Model Detail Views

4.9 Model Detail Views

Model detail views are algorithm-specific. Viewing the model detail views will provide you with additional information about the model you created. The names of model detail views begin with DM$. Some model views, such as Global Name-Value Pairs view (DM$VGmodel_name), Computed Settings view (DM$VSmodel_name), Model Build Alerts view (DM$VWmodel_name), and Normalization and Missing Value Handling view (DM$VNmodel_name), are shared by all algorithms and are documented separately. Aside from that, classification, clustering, and regression algorithms share some common views. The columns returned by these views may differ between algorithms.

The following are the model views:

Model Detail Views for Association Rules
The model detail view DM$VRmodel_name contains the generated rules for association models.
Model Detail View for Frequent Itemsets
The model detail view DM$VImodel_name contains information about frequent itemsets.
Model Detail Views for Transactional Itemsets
The model detail view DM$VTmodel_name contains information about the transactional itemsets.
Model Detail View for Transactional Rule
The model detail view DM$VAmodel_name contains information about transactional rules and transactional itemsets.
Model Detail Views for Classification Algorithms
Model detail views for classification algorithms are the target map view and scoring cost view, which are applicable to all classification algorithms.
Model Detail Views for CUR Matrix Decomposition
Model detail views for CUR Matrix Decomposition contain information about the scores and ranks of attributes and rows.
Model Detail Views for Decision Tree
The model detail views specific to Decision Tree are the hierarchy view, node statistics view, node description view, and the cost matrix view.
Model Detail Views for Generalized Linear Model
Model detail views specific to Generalized Linear Model (GLM) such as details and row diagnostics for linear and logistic regression models are discussed.
Model Detail View for Multivariate State Estimation Technique - Sequential Probability Ratio Test
The model detail view specific to Multivariate State Estimation Technique - Sequential Probability Ratio Test contains information about Global Name-Value Paris.
Model Detail Views for Naive Bayes
The model detail views specific to Naive Bayes are the prior view and result view.
Model Detail Views for Neural Network
Model detail views specific to Neural Network contain information about the weights of the neurons: input layer and hidden layers.
Model Detail Views for Random Forest
Model detail views specific to Random Forest contain variable importance measures and statistics.
Model Detail View for Support Vector Machine
Model detail views specific to Support Vector Machine (SVM) contain linear coefficients and support vector statistics.
Model Detail Views for XGBoost
The model detail views specific to XGBoost contain information about Feature Importance view and Global Name-Value Pairs view.
Model Detail Views for Clustering Algorithms
Oracle Machine Learning for SQL supports these clustering algorithms: Expectation Maximization (EM), k-Means (KM), and orthogonal partitioning clustering (O-Cluster, OC).
Model Detail Views for Expectation Maximization
Model detail views specific to Expectation Maximization (EM) contain additional information about an EM model. Additional views are available for EM Clustering, but are absent for EM Anomaly.
Model Detail Views for k-Means
Model detail views specific to k-Means (KM) contain clustering description view (DM$VG), and scoring information.
Model Detail Views for O-Cluster
Model detail views specific to O-Cluster (OC) contain information about description view, histograms view, and global view.
Model Detail Views for Explicit Semantic Analysis
Model detail views specific to Explicit Semantic Analysis (ESA) contain information about attribute statistics and features.
Model Detail Views for Non-Negative Matrix Factorization
Model detail views specific to Non-Negative Matrix Factorization (NMF) contain information about the encoding H matrix and H inverse matrix.
Model Detail Views for Singular Value Decomposition
Model detail views specific to Singular Value Decomposition (SVD) contain information about the S matrix, right-singular vectors, and left-singular vectors.
Model Detail Views for Minimum Description Length
Model detail views specific to Minimum Description Length (MDL) (for calculating attribute importance) contain information about attribute importance models.
Model Detail Views for Binning
The binning view DM$VB describes the bin boundaries used in automatic data preparation.
Model Detail Views for Global Information
Model detail views for global information contain information about global statistics, alerts, and computed settings.
Model Detail Views for Normalization and Missing Value Handling
The Normalization and Missing Value Handling view DM$VN describes the normalization parameters used in Automatic Data Preparation (ADP) and the missing value replacement when a NULL value is encountered. Missing value replacement applies only to the twodimensional columns and does not apply to the nested columns.
Model Detail Views for Exponential Smoothing
Model detail views specific to Exponential Smoothing (ESM) include information about the model output, global information about the model, and views that support time series regression.
Model Detail Views for Text Features
The model details view for text features is DM$VXmodel_name.
Model Detail Views for ONNX Models
You can view the details of an embedding model using the model detail views. The names of the views begin with DM$V.

Parent topic: Create a Model

4.9.1 Model Detail Views for Association Rules

The model detail view DM$VRmodel_name contains the generated rules for association models.

These are the available model views for Association Rules:

Model Views	Description
`DM$VAmodel_name`	Association Rules For Transactional Data
`DM$VG`model_name	Global Name-Value Pairs
`DM$VI`model_name:	Association Rule Itemsets
`DM$VR`model_name	Association Rules
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Association Rule Itemsets For Transactional Data
`DM$VW`model_name	Model Build Alerts

Depending on the settings of the model, this rule view (DM$VRmodel_name) different sets of columns. Settings ODMS_ITEM_ID_COLUMN_NAME and ODMS_ITEM_VALUE_COLUMN_NAME determine how each item is defined. If ODMS_ITEM_ID_COLUMN_NAME is set, the input format is called transactional input, otherwise, the input format is called 2-Dimensional input. With transactional input, if setting ODMS_ITEM_VALUE_COLUMN_NAME is not set, each item is defined by ITEM_NAME, otherwise, each item is defined by ITEM_NAME and ITEM_VALUE. With 2-Dimensional input, each item is defined by ITEM_NAME, ITEM_SUBNAME and ITEM_VALUE. Setting ASSO_AGGREGATES specifies the columns to aggregate, which is displayed in the view.

Note:

Setting ASSO_AGGREGATES is not allowed for 2-dimensional input.

The following shows the views with different settings.

Transactional Input Without ASSO_AGGREGATES Setting

When you sett ITEM_NAME (ODMS_ITEM_ID_COLUMN_NAME) and do not set ITEM_VALUE (ODMS_ITEM_VALUE_COLUMN_NAME), the view contains the following. The consequent item is defined with only the name field. If you also set ITEM_VALUE, the view has the additional column CONSEQUENT_VALUE that specifies the value field.

Name                                      Type
 ----------------------------------------- ----------------------------
 PARTITION_NAME                            VARCHAR2(128)
 RULE_ID                                   NUMBER
 RULE_SUPPORT                              NUMBER
 RULE_CONFIDENCE                           NUMBER
 RULE_LIFT                                 NUMBER
 RULE_REVCONFIDENCE                        NUMBER
 ANTECEDENT_SUPPORT                        NUMBER
 NUMBER_OF_ITEMS                           NUMBER
 CONSEQUENT_SUPPORT                        NUMBER
 CONSEQUENT_NAME                           VARCHAR2(4000)
 ANTECEDENT                                SYS.XMLTYPE

Table 4-16 Rule View Columns for Transactional Inputs

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model to retrieve details.
`RULE_ID`	The identifier of the rule.
`RULE_SUPPORT`	The number of transactions that satisfy the rule.
`RULE_CONFIDENCE`	The likelihood of a transaction satisfying the rule.
`RULE_LIFT`	The degree of improvement in the prediction over random chance when the rule is satisfied.
`RULE_REVCONFIDENCE`	The number of transactions in which the rule occurs divided by the number of transactions in which the consequent occurs.
`ANTECEDENT_SUPPORT`	The ratio of the number of transactions that satisfy the antecedent to the total number of transactions.
`NUMBER_OF_ITEMS`	The total number of attributes referenced in the antecedent and consequent of the rule.
`CONSEQUENT_SUPPORT`	The ratio of the number of transactions that satisfy the consequent to the total number of transactions.
`CONSEQUENT_NAME`	The name of the consequent.
`CONSEQUENT_VALUE`	The value of the consequent. This column is present when `Item_value` (`ODMS_ITEM_VALUE_COLUMN_NAME`) is set with `TYPE` as numerical or categorical.
`ANTECEDENT`	The antecedent is described as an itemset. At the itemset level, it specifies the number of aggregates, and if not zero, the names of the columns to be aggregated (as well as the mapping to `ASSO_AGG`*). The itemset contains `>= 1` items. When `ODMS_ITEM_VALUE_COLUMN_NAME` is not set, each item is defined by `item_name`. As an example, if the antecedent contains one item B, then it is represented as follows: `<itemset NUMAGGR="0"><item><item_name>B</item_name></item></itemset>` As another example, if the antecedent contains two items, A and C, then it is represented as follows: `<itemset NUMAGGR="0"><item><item_name>A</item_name></item><item><item_name>C</item_name></item></itemset>` When setting `ODMS_ITEM_VALUE_COLUMN_NAME` is set, each item is defined by `item_name` and `item_value`. As an example, if the antecedent contains two items, (name A, value 1) and (name C, value 1), then it is represented as follows: `<itemset NUMAGGR="0"><item><item_name>A</item_name><item_value>1</item_value></item><item><item_name>C</item_name><item_value>1</item_value></item></itemset>`

Transactional Input With ASSO_AGGREGATES Setting

Similar to the view without an aggregates setting, there are three cases:

Rule view when ODMS_ITEM_ID_COLUMN_NAME is set and Item_value (ODMS_ITEM_VALUE_COLUMN_NAME) is not set.
Rule view when ODMS_ITEM_ID_COLUMN_NAME is set and Item_value (ODMS_ITEM_VALUE_COLUMN_NAME) is set with TYPE as numerical, the view has a CONSEQUENT_VALUE column.
Rule view when ODMS_ITEM_ID_COLUMN_NAME is set and Item_value (ODMS_ITEM_VALUE_COLUMN_NAME) is set with TYPE as categorical, the view has a CONSEQUENT_VALUE column.

For the example that produces the following rules, see “Example: Calculating Aggregates” in Oracle Machine Learning for SQL Concepts.

The view reports two sets of aggregates results:

ANT_RULE_PROFIT refers to the total profit for the antecedent itemset with respect to the rule, the profit for each individual item of the antecedent itemset is shown in the ANTECEDENT(XMLtype) column, CON_RULE_PROFIT refers to the total profit for the consequent item with respect to the rule.

In the example, for rule (A, B) => C, the rule itemset (A, B, C) occurs in the transactions of customer 1 and customer 3. The ANT_RULE_PROFIT is $21.20, The ANTECEDENT is shown as follow, which tells that item A has profit 5.00 + 3.00 = $8.00 and item B has profit 3.20 + 10.00 = $13.20, which sum up to ANT_RULE_PROFIT.
```
<itemset NUMAGGR="1" ASSO_AGG0="profit"><item><item_name>A</item_name><ASSO_AGG0>8.0E+000</ASSO_AGG0></item><item><item_name>B</item_name><ASSO_AGG0>1.32E+001</ASSO_AGG0></item></itemset> 
The CON_RULE_PROFIT is 12.00 + 14.00 = $26.00
```
ANT_PROFIT refers to the total profit for the antecedent itemset, while CON_PROFIT refers to the total profit for the consequent item. The difference between CON_PROFIT and CON_RULE_PROFIT (the same applies to ANT_PROFIT and ANT_RULE_PROFIT) is that CON_PROFIT counts all profit for the consequent item across all transactions where the consequent occurs, while CON_RULE_PROFIT only counts across transactions where the rule itemset occurs.

For example, item C occurs in transactions for customer 1, 2 and 3, CON_PROFIT is 12.00 + 4.20 + 14.00 = $30.20, while CON_RULE_PROFIT only counts transactions for customer 1 and 3 where the rule itemset (A, B, C) occurs.

Similarly, ANT_PROFIT counts all transactions where itemset (A, B) occurs, while ANT_RULE_PROFIT counts only transactions where the rule itemset (A, B, C) occurs. In this example, by coincidence, both count transactions for customer 1 and 3, and have the same value.

Example 4-16 Examples

The following example shows the view when setting ASSO_AGGREGATES specifies column profit and column sales to be aggregated. In this example, ITEM_VALUE column is not specified.

Name                                      Type
 ----------------------------------------- ----------------------------
 PARTITION_NAME                            VARCHAR2(128)
 RULE_ID                                   NUMBER
 RULE_SUPPORT                              NUMBER
 RULE_CONFIDENCE                           NUMBER
 RULE_LIFT                                 NUMBER
 RULE_REVCONFIDENCE                        NUMBER
 ANTECEDENT_SUPPORT                        NUMBER
 NUMBER_OF_ITEMS                           NUMBER
 CONSEQUENT_SUPPORT                        NUMBER
 CONSEQUENT_NAME                           VARCHAR2(4000)
 ANTECEDENT                                SYS.XMLTYPE
 ANT_RULE_PROFIT                           BINARY_DOUBLE
 CON_RULE_PROFIT                           BINARY_DOUBLE
 ANT_PROFIT                                BINARY_DOUBLE
 CON_PROFIT                                BINARY_DOUBLE
 ANT_RULE_SALES                            BINARY_DOUBLE
 CON_RULE_SALES                            BINARY_DOUBLE
 ANT_SALES                                 BINARY_DOUBLE
 CON_SALES                                 BINARY_DOUBLE

The rule view has a CONSEQUENT_VALUE column when ODMS_ITEM_ID_COLUMN_NAME is set and Item_value (ODMS_ITEM_VALUE_COLUMN_NAME) is set with TYPE as numerical or categorical.

2-Dimensional Inputs

In Oracle Machine Learning for SQL, association models can be built using either transactional or two-dimensional data formats. For two-dimensional input, each item is defined by three fields: NAME, VALUE and SUBNAME. The NAME field is the name of the column. The VALUE field is the content of the column. The SUBNAME field is used when the input data table contains a nested table. In that case, SUBNAME is the name of the nested table's column. See, Example: Creating a Nested Column for Market Basket Analysis. In this example, there is a nested column. The CONSEQUENT_SUBNAME is the ATTRIBUTE_NAME part of the nested column. That is, 'O/S Documentation Set - English' and CONSEQUENT_VALUE is the value part of the nested column, which is, 1.

The view uses three columns for the consequent. The rule view has the following columns:

Name                                      Type
 ----------------------- --------------------- 
 PARTITION_NAME                            VARCHAR2(128)
 RULE_ID                                   NUMBER
 RULE_SUPPORT                              NUMBER
 RULE_CONFIDENCE                           NUMBER
 RULE_LIFT                                 NUMBER
 RULE_REVCONFIDENCE                        NUMBER
 ANTECEDENT_SUPPORT                        NUMBER
 NUMBER_OF_ITEMS                           NUMBER
 CONSEQUENT_SUPPORT                        NUMBER
 CONSEQUENT_NAME                           VARCHAR2(4000)
 CONSEQUENT_SUBNAME                        VARCHAR2(4000)
 CONSEQUENT_VALUE                          VARCHAR2(4000)
 ANTECEDENT                                SYS.XMLTYPE

Note:

All of the types for three columns for the consequent are VARCHAR2. ASSO_AGGREGATES is not applicable for 2-Dimensional input format.

The following table displays rule view columns for 2-Dimensional input with the descriptions of only the fields that are specific to 2-D inputs.

Table 4-17 Rule View for 2-Dimensional Input

Column Name Description

Column Name	Description
`CONSEQUENT_SUBNAME`	For two-dimensional inputs, `CONSEQUENT_SUBNAME` is used for nested column in the input data table.
`CONSEQUENT_VALUE`	The value of the consequent when setting `Item_value` is set with `TYPE` as numerical or categorical.
`ANTECEDENT`	The antecedent is described as an itemset. The itemset contains `>= 1` items. Each item is defined using `ITEM_NAME`, `ITEM_SUBNAME`, and `ITEM_VALUE`: As an example, assuming that this is not a nested table input, and the antecedent contains one item: (name `ADDR`, value `MA`). The antecedent (XMLtype) is as follows: `<itemset NUMAGGR="0"><item><item_name>ADDR</item_name><item_subname></item_subna me><item_value>MA</item_value></item></itemset>` For 2-Dimensional input with nested table, the subname field is filled.

CONSEQUENT_SUBNAME

For two-dimensional inputs, CONSEQUENT_SUBNAME is used for nested column in the input data table.

CONSEQUENT_VALUE

The value of the consequent when setting Item_value is set with TYPE as numerical or categorical.

ANTECEDENT

The antecedent is described as an itemset. The itemset contains >= 1 items. Each item is defined using ITEM_NAME, ITEM_SUBNAME, and ITEM_VALUE:

As an example, assuming that this is not a nested table input, and the antecedent contains one item: (name ADDR, value MA). The antecedent (XMLtype) is as follows:

<itemset NUMAGGR="0"><item><item_name>ADDR</item_name><item_subname></item_subna
me><item_value>MA</item_value></item></itemset>

For 2-Dimensional input with nested table, the subname field is filled.

Global Name-Value Pairs View for Association Rules

Global Name-Value Pairs View produces a single column for an association model. The following table describes the columns returned for association model.

Table 4-18 Global Name-Value Pairs View for an Association Model

Name	Description
`ITEMSET_COUNT`	The number of itemsets generated.
`MAX_SUPPORT`	The maximum support.
`NUM_ROWS`	The total number of rows used in the build.
`RULE_COUNT`	The number of association rules in the model generated.
`TRANSACTION_COUNT`	The number of the transactions in the input data.

Parent topic: Model Detail Views

4.9.2 Model Detail View for Frequent Itemsets

The model detail view DM$VImodel_name contains information about frequent itemsets.

The Association Rule Itemsets view (DM$VImodel_name) has the following columns:

Name               Type
-------------      ------------------
PARTITION_NAME     VARCHAR2 (128)
ITEMSET_ID         NUMBER
SUPPORT            NUMBER
NUMBER_OF_ITEMS    NUMBER
 ITEMSET            SYS.XMLTYPE

Table 4-19 Association Rule Itemsets View

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model
`ITEMSET_ID`	Itemset identifier
`SUPPORT`	Support of the itemset
`NUMBER_OF_ITEMS`	Number of items in the itemset
`ITEMSET`	Frequent itemset The structure of the `SYS.XMLTYPE` column itemset is the same as the corresponding Antecedent column of the rule view.

Parent topic: Model Detail Views

4.9.3 Model Detail Views for Transactional Itemsets

The model detail view DM$VTmodel_name contains information about the transactional itemsets.

For the very common case of transactional data without aggregates, the Association Rule Itemsets For Transactional Data view (DM$VTmodel_name) provides the itemsets information in transactional format. This view can help improve performance for some queries as compared to the view with the XML column. The transactional itemsets view has the following columns:

Name                Type
-----------------   -----------------
PARTITION_NAME      VARCHAR2(128)
ITEMSET_ID          NUMBER
ITEM_ID             NUMBER
SUPPORT             NUMBER
NUMBER_OF_ITEMS     NUMBER
ITEM_NAME           VARCHAR2(4000)

Table 4-20 Association Rule Itemsets For Transactional Data View

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model
`ITEMSET_ID`	Itemset identifier
`ITEM_ID`	Item identifier
`SUPPORT`	Support of the itemset
`NUMBER_OF_ITEMS`	Number of items in the itemset
`ITEM_NAME`	The name of the item

Parent topic: Model Detail Views

4.9.4 Model Detail View for Transactional Rule

The model detail view DM$VAmodel_name contains information about transactional rules and transactional itemsets.

Transactional data without aggregates also has an Association Rules For Transactional Data view (DM$VAmodel_name). This view can improve performance for some queries as compared to the view with the XML column. The transactional rule view has the following columns:

Name                                      Type
----------------------------------------- ----------------------------
PARTITION_NAME                            VARCHAR2(128)
RULE_ID                                   NUMBER
ANTECEDENT_PREDICATE                      VARCHAR2(4000)
CONSEQUENT_PREDICATE                      VARCHAR2(4000)
RULE_SUPPORT                              NUMBER
RULE_CONFIDENCE                           NUMBER
RULE_LIFT                                 NUMBER
RULE_REVCONFIDENCE                        NUMBER
RULE_ITEMSET_ID                           NUMBER
ANTECEDENT_SUPPORT                        NUMBER
CONSEQUENT_SUPPORT                        NUMBER
NUMBER_OF_ITEMS                           NUMBER

Table 4-21 Association Rules For Transactional Data View

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model
`RULE_ID`	Rule identifier
`ANTECEDENT_PREDICATE`	Name of the Antecedent item.
`CONSEQUENT_PREDICATE`	Name of the Consequent item
`RULE_SUPPORT`	Support of the rule
`RULE_CONFIDENCE`	The likelihood a transaction satisfies the rule when it contains the Antecedent.
`RULE_LIFT`	The degree of improvement in the prediction over random chance when the rule is satisfied
`RULE_REVCONFIDENCE`	The number of transactions in which the rule occurs divided by the number of transactions in which the consequent occurs
`RULE_ITEMSET_ID`	Itemset identifier
`ANTECEDENT_SUPPORT`	The ratio of the number of transactions that satisfy the antecedent to the total number of transactions
`CONSEQUENT_SUPPORT`	The ratio of the number of transactions that satisfy the consequent to the total number of transactions
`NUMBER_OF_ITEMS`	Number of items in the rule

Parent topic: Model Detail Views

4.9.5 Model Detail Views for Classification Algorithms

Model detail views for classification algorithms are the target map view and scoring cost view, which are applicable to all classification algorithms.

These are the available model views for Classification algorithm:

Model Views	Description
`DM$VA`model_name	Variable Importance
`DM$VC`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VW`model_name:	Model Build Alerts

The Classification Targets view (DM$VTmodel_name) describes the target distribution for classification models. The view has the following columns:

Name                                      Type
----------------------------------------- ----------------------------
PARTITION_NAME                            VARCHAR2(128)
TARGET_VALUE                              NUMBER/VARCHAR2 
TARGET_COUNT                              NUMBER
TARGET_WEIGHT                             NUMBER

Table 4-22 Classification Targets View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`TARGET_VALUE`	Target value, numerical or categorical
`TARGET_COUNT`	Number of rows for a given `TARGET_VALUE`
`TARGET_WEIGHT`	Weight for a given `TARGET_VALUE`

The Scoring Cost Matrix view (DM$VCmodel_name) describes the scoring cost matrix for classification models. The view has the following columns:

Name                                      Type
----------------------------------------- --------------------------------
PARTITION_NAME                            VARCHAR2(128)
ACTUAL_TARGET_VALUE                       NUMBER/VARCHAR2  
PREDICTED_TARGET_VALUE                    NUMBER/VARCHAR2  
COST                                      NUMBER

Table 4-23 Scoring Cost Matrix View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ACTUAL_TARGET_VALUE`	A valid target value
`PREDICTED_TARGET_VALUE`	Predicted target value
`COST`	Associated cost for the actual and predicted target value pair

Parent topic: Model Detail Views

4.9.6 Model Detail Views for CUR Matrix Decomposition

Model detail views for CUR Matrix Decomposition contain information about the scores and ranks of attributes and rows.

CUR Matrix Decomposition models have the following views:

Attribute importance and rank: DM$VCmodel_name

Row importance and rank: DM$VRmodel_name

Global statistics: DM$VG

The attribute importance and rank view DM$VCmodel_name has the following columns:

Name                           Type               
-----------------              -----------------      
PARTITION_NAME                 VARCHAR2(128)       
ATTRIBUTE_NAME                 VARCHAR2(128)       
ATTRIBUTE_SUBNAME              VARCHAR2(4000)      
ATTRIBUTE_VALUE                VARCHAR2(4000)    
ATTRIBUTE_IMPORTANCE           NUMBER     
ATTRIBUTE_RANK                 NUMBER

Table 4-24 Attribute Importance and Rank View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ATTRIBUTE_NAME`	Attribute name
`ATTRIBUTE_SUBNAME`	Attribute subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Value of the attribute
`ATTRIBUTE_IMPORTANCE`	Attribute leverage score
`ATTRIBUTE_RANK`	Attribute rank based on leverage score

The view DM$VRmodel_name exposes the leverage scores and ranks of all selected rows through a view. This view is created when users decide to perform row importance and the CASE_ID column is present. The view has the following columns:

Name	               Type		                   
---------------------     ------------------------	   
PARTITION_NAME            VARCHAR2(128)         	   
CASE_ID                   Original cid data types,         
                          including NUMBER, VARCHAR2, 
                          DATE, TIMESTAMP, 
                          TIMESTAMP WITH TIME ZONE, 
                          TIMESTAMP WITH LOCAL TIME ZONE	
ROW_IMPORTANCE            NUMBER	                       
ROW_RANK                  NUMBER

Table 4-25 Row Importance and Rank View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`CASE_ID`	Case ID. The supported case ID types are the same as that supported for GLM, SVD, and ESA algorithms.
`ROW_IMPORTANCE`	Row leverage score
`ROW_RANK`	Row rank based on leverage score

The following table describes global statistics for CUR Matrix Decomposition.

Table 4-26 CUR Matrix Decomposition Statistics Information In Model Global View.

Name	Description
`NUM_COMPONENTS`	Number of SVD components (SVD rank)
`NUM_ROWS`	Number of rows used in the model build

Parent topic: Model Detail Views

4.9.7 Model Detail Views for Decision Tree

The model detail views specific to Decision Tree are the hierarchy view, node statistics view, node description view, and the cost matrix view.

These are the model views available for Decision Tree:

Model Views	Description
`DM$VC`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VI`model_name	Decision Tree Statistics
`DM$VM`model_name	Decision Tree Build Cost Matrix
`DM$VO`model_name	Decision Tree Nodes
`DM$VP`model_name	Decision Tree Hierarchy
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VW`model_name	Model Build Alerts

The Decision Tree Hierarchy view (DM$VPmodel_name) describes the decision tree hierarchy and the split information for each level in the decision tree. The view has the following columns:

Name                               Type
---------------------------------- ---------------------------
PARTITION_NAME                     VARCHAR2(128)
PARENT                             NUMBER              
SPLIT_TYPE                         VARCHAR2            
NODE                               NUMBER              
ATTRIBUTE_NAME                     VARCHAR2(128)       
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)      
OPERATOR                           VARCHAR2            
VALUE                              SYS.XMLTYPE

Table 4-27 Decision Tree Hierarchy View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`PARENT`	Node ID of the parent
`SPLIT_TYPE`	The main or surrogate split
`NODE`	The node ID
`ATTRIBUTE_NAME`	The attribute used as the splitting criterion at the parent node to produce this node.
`ATTRIBUTE_SUBNAME`	Split attribute subname. The value is null for non-nested columns.
`OPERATOR`	Split operator
`VALUE`	Value used as the splitting criterion. This is an XML element described using the `<Element>` tag. For example, `<Element>Windy</Element><Element>Hot</Element>`.

The Decision Tree Statistics view (DM$VImodel_name) describes the statistics associated with individual tree nodes. The statistics include a target histogram for the data in the node. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
NODE                               NUMBER      
NODE_SUPPORT                       NUMBER       
PREDICTED_TARGET_VALUE             NUMBER/VARCHAR2
TARGET_VALUE                       NUMBER/VARCHAR2 
TARGET_SUPPORT                     NUMBER

Table 4-28 Decision Tree Statistics View

Parameter	Description
`PARTITION_NAME`	Partition name in a partitioned model
`NODE`	The node ID
`NODE_SUPPORT`	Number of records in the training set that belong to the node
`PREDICTED_TARGET_VALUE`	Predicted Target value
`TARGET_VALUE`	A target value seen in the training data
`TARGET_SUPPORT`	The number of records that belong to the node and have the value specified in the `TARGET_VALUE` column

The Decision Tree Nodes (DM$VOmodel_name) view describes higher level node. The DM$VOmodel_name has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
NODE                               NUMBER                
NODE_SUPPORT                       NUMBER                
PREDICTED_TARGET_VALUE             NUMBER/VARCHAR2       
PARENT                             NUMBER               
ATTRIBUTE_NAME                     VARCHAR2(128)       
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)       
OPERATOR                           VARCHAR2             
VALUE                              SYS.XMLTYPE

Table 4-29 Decision Tree Nodes View

Parameter	Description
`PARTITION_NAME`	Partition name in a partitioned model
`NODE`	The node ID
`NODE_SUPPORT`	Number of records in the training set that belong to the node
`PREDICTED_TARGET_VALUE`	Predicted Target value
`PARENT`	The ID of the parent
`ATTRIBUTE_NAME`	Specifies the attribute name
`ATTRIBUTE_SUBNAME`	Specifies the attribute subname
`OPERATOR`	Attribute predicate operator - a conditional operator taking the following values: IN, = , <>, < , >, <=, and >=
`VALUE`	Value used as the description criterion. This is an XML element described using the `<Element>` tag. For example, `<Element>Windy</Element><Element>Hot</Element>`.

The Decision Tree Build Cost Matrix view (DM$VMmodel_name) describes the cost matrix used by the Decision Tree build. The DM$VMmodel_name view has the following columns:

Name                            Type
----------------------------------------- --------------------------------
PARTITION_NAME                            VARCHAR2(128)
ACTUAL_TARGET_VALUE                       NUMBER/VARCHAR2 
PREDICTED_TARGET_VALUE                    NUMBER/VARCHAR2 
COST                                      NUMBER

Table 4-30 Decision Tree Build Cost Matrix View

Parameter	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ACTUAL_TARGET_VALUE`	Valid target value
`PREDICTED_TARGET_VALUE`	Predicted Target value
`COST`	Associated cost for the actual and predicted target value pair

The following table describes the Global Name-Value Pairs view (DM$VGmodel_name) columns specific to a Decision Tree model.

Table 4-31 Global Name-Value Pairs View

Name	Description
`NUM_ROWS`	The total number of rows used in the build

Parent topic: Model Detail Views

4.9.8 Model Detail Views for Generalized Linear Model

Model detail views specific to Generalized Linear Model (GLM) such as details and row diagnostics for linear and logistic regression models are discussed.

The following model views are available for GLM:

Model Views	Description
`DM$VA`model_name	GLM Regression Row Diagnostics
`DM$VD`model_name	GLM Regression Attribute Diagnostics
`DM$VG`model_name	Global Name-Value Pairs
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts

The GLM Regression Attribute Diagnostics view (DM$VDmodel_name) describes the final model information for both linear regression models and logistic regression models.

For linear regression, the view DM$VDmodel_name has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
ATTRIBUTE_VALUE                    VARCHAR2(4000)
FEATURE_EXPRESSION                 VARCHAR2(4000)
COEFFICIENT                        BINARY_DOUBLE
STD_ERROR                          BINARY_DOUBLE
TEST_STATISTIC                     BINARY_DOUBLE
P_VALUE                            BINARY_DOUBLE
VIF                                BINARY_DOUBLE
STD_COEFFICIENT                    BINARY_DOUBLE
LOWER_COEFF_LIMIT                  BINARY_DOUBLE
UPPER_COEFF_LIMIT                  BINARY_DOUBLE

For logistic regression, the view DM$VDmodel_name has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
TARGET_VALUE                       NUMBER/VARCHAR2
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
ATTRIBUTE_VALUE                    VARCHAR2(4000)
FEATURE_EXPRESSION                 VARCHAR2(4000)
COEFFICIENT                        BINARY_DOUBLE
STD_ERROR                          BINARY_DOUBLE
TEST_STATISTIC                     BINARY_DOUBLE
P_VALUE                            BINARY_DOUBLE
STD_COEFFICIENT                    BINARY_DOUBLE
LOWER_COEFF_LIMIT                  BINARY_DOUBLE
UPPER_COEFF_LIMIT                  BINARY_DOUBLE
EXP_COEFFICIENT                    BINARY_DOUBLE
EXP_LOWER_COEFF_LIMIT              BINARY_DOUBLE
EXP_UPPER_COEFF_LIMIT              BINARY_DOUBLE

Table 4-32 Model View for Linear and Logistic Regression Models

Column Name	Description
`PARTITION_NAME`	The name of a feature in the model
`TARGET_VALUE`	Valid target value
`ATTRIBUTE_NAME`	The attribute name when there is no subname, or first part of the attribute name when there is a subname. `ATTRIBUTE_NAME` is the name of a column in the source table or view. If the column is a non-nested, numeric column, then `ATTRIBUTE_NAME` is the name of the machine learning attribute. For the intercept, `ATTRIBUTE_NAME` is null. Intercepts are equivalent to the bias term in SVM models.
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns. When the nested column is numeric, the machine learning attribute is identified by the combination `ATTRIBUTE_NAME - ATTRIBUTE_SUBNAME`. If the column is not nested, `ATTRIBUTE_SUBNAME` is null. If the attribute is an intercept, both the `ATTRIBUTE_NAME` and the `ATTRIBUTE_SUBNAME` are null.
`ATTRIBUTE_VALUE`	A unique value that can be assumed by a categorical column or nested categorical column. For categorical columns, a machine learning attribute is identified by a unique `ATTRIBUTE_NAME.ATTRIBUTE_VALUE` pair. For nested categorical columns, a machine learning attribute is identified by the combination: `ATTRIBUTE_NAME.ATTRIBUTE_SUBNAME.ATTRIBUTE_VALUE`. For numerical attributes, `ATTRIBUTE_VALUE` is null.
`FEATURE_EXPRESSION`	The feature name constructed by the algorithm when feature selection is enabled. If feature selection is not enabled, the feature name is the fully-qualified attribute name (attribute_name.attribute_subname if the attribute is in a nested column). For categorical attributes, the algorithm constructs a feature name that has the following form: fully-qualified_attribute_name.attribute_value When feature generation is enabled, a term in the model can be a single machine learning attribute or the product of up to 3 machine learning attributes. Component machine learning attributes can be repeated within a single term. If feature generation is not enabled or, if feature generation is enabled, but no multiple component terms are discovered by the `CREATE` model process, then `FEATURE_EXPRESSION` is null. Note: In 12c Release 2, the algorithm does not subtract the mean from numerical components.
`COEFFICIENT`	The estimated coefficient.
`STD_ERROR`	Standard error of the coefficient estimate.
`TEST_STATISTIC`	For linear regression, the t-value of the coefficient estimate. For logistic regression, the Wald chi-square value of the coefficient estimate.
`P_VALUE`	Probability of the `TEST_STATISTIC` under the (NULL) hypothesis that the term in the model is not statistically significant. A low probability indicates that the term is significant, while a high probability indicates that the term can be better discarded. Used to analyze the significance of specific attributes in the model.
`VIF`	Variance Inflation Factor. The value is zero for the intercept. For logistic regression, `VIF` is null.
`STD_COEFFICIENT`	Standardized estimate of the coefficient.
`LOWER_COEFF_LIMIT`	Lower confidence bound of the coefficient.
`UPPER_COEFF_LIMIT`	Upper confidence bound of the coefficient.
`EXP_COEFFICIENT`	Exponentiated coefficient for logistic regression. For linear regression, `EXP_COEFFICIENT` is null.
`EXP_LOWER_COEFF_LIMIT`	Exponentiated coefficient for lower confidence bound of the coefficient for logistic regression. For linear regression, `EXP_LOWER_COEFF_LIMIT` is null.
`EXP_UPPER_COEFF_LIMIT`	Exponentiated coefficient for upper confidence bound of the coefficient for logistic regression. For linear regression, `EXP_UPPER_COEFF_LIMIT` is null.

The GLM Regression Row Diagnostics view DM$VAmodel_name describes row level information for both linear regression models and logistic regression models. For linear regression, the view DM$VAmodel_name has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
CASE_ID                            NUMBER/VARHCAR2, DATE, TIMESTAMP, 
                                   TIMESTAMP WITH TIME ZONE,
                                   TIMESTAMP WITH LOCAL TIME ZONE
TARGET_VALUE                       BINARY_DOUBLE
PREDICTED_TARGET_VALUE             BINARY_DOUBLE
Hat                                BINARY_DOUBLE
RESIDUAL                           BINARY_DOUBLE
STD_ERR_RESIDUAL                   BINARY_DOUBLE
STUDENTIZED_RESIDUAL               BINARY_DOUBLE
PRED_RES                           BINARY_DOUBLE
COOKS_D                            BINARY_DOUBLE

Table 4-33 GLM Regression Row Diagnostics View for Linear Regression

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`CASE_ID`	Name of the case identifier
`TARGET_VALUE`	The actual target value as taken from the input row
`PREDICTED_TARGET_VALUE`	The model predicted target value for the row
`HAT`	The diagonal element of the n*n (n=number of rows) that the Hat matrix identifies with a specific input row. The model predictions for the input data are the product of the Hat matrix and vector of input target values. The diagonal elements (Hat values) represent the influence of the i^th row on the i^th fitted value. Large Hat values are indicators that the i^th row is a point of high leverage, a potential outlier.
`RESIDUAL`	The difference between the predicted and actual target value for a specific input row.
`STD_ERR_RESIDUAL`	The standard error residual, sometimes called the Studentized residual, re-scales the residual to have constant variance across all input rows in an effort to make the input row residuals comparable. The process multiplies the residual by square root of the row weight divided by the product of the model mean square error and 1 minus the Hat value.
`STUDENTIZED_RESIDUAL`	Studentized deletion residual adjusts the standard error residual for the influence of the current row.
`PRED_RES`	The predictive residual is the weighted square of the deletion residuals, computed as the row weight multiplied by the square of the residual divided by 1 minus the Hat value.
`COOKS_D`	Cook's distance is a measure of the combined impact of the i^th case on all of the estimated regression coefficients.

For logistic regression, the view DM$VAmodel_name has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
CASE_ID                            NUMBER/VARHCAR2, DATE, TIMESTAMP, 
                                   TIMESTAMP WITH TIME ZONE,
                                   TIMESTAMP WITH LOCAL TIME ZONE
TARGET_VALUE                       NUMBER/VARCHAR2
TARGET_VALUE_PROB                  BINARY_DOUBLE
Hat                                BINARY_DOUBLE
WORKING_RESIDUAL                   BINARY_DOUBLE
PEARSON_RESIDUAL                   BINARY_DOUBLE
DEVIANCE_RESIDUAL                  BINARY_DOUBLE
C                                  BINARY_DOUBLE
CBAR                        	    BINARY_DOUBLE
DIFDEV                             BINARY_DOUBLE
DIFCHISQ                           BINARY_DOUBLE

Table 4-34 GLM Regression Row Diagnostics View for Logistic Regression

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`CASE_ID`	Name of the case identifier
`TARGET_VALUE`	The actual target value as taken from the input row
`TARGET_VALUE_PROB`	Model estimate of the probability of the predicted target value.
`Hat`	The Hat value concept from linear regression is extended to logistic regression by multiplying the linear regression Hat value by the variance function for logistic regression, the predicted probability multiplied by 1 minus the predicted probability.
`WORKING_RESIDUAL`	The working residual is the residual of the working response. The working response is the response on the linearized scale. For logistic regression it has the form: the i^th row residual divided by the variance of the i^th row prediction. The variance of the prediction is the predicted probability multiplied by 1 minus the predicted probability. `WORKING_RESIDUAL` is the difference between the working response and the linear predictor at convergence.
`PEARSON_RESIDUAL`	The Pearson residual is a re-scaled version of the working residual, accounting for the weight. For logistic regression, the Pearson residual multiplies the residual by a factor that is computed as square root of the weight divided by the variance of the predicted probability for the i^th row. `RESIDUAL` is 1 minus the predicted probability of the actual target value for the row.
`DEVIANCE_RESIDUAL`	The `DEVIANCE_RESIDUAL` is the contribution to the model deviance of the i^th observation. For logistic regression it has the form the square root of 2 times the `log(1 + e^eta) - eta` for the non-reference class and -square root of 2 time the `log (1 + eta)` for the reference class, where `eta` is the linear prediction (the prediction as if the model were a linear regression).
`C`	Measures the overall change in the fitted logits due to the deletion of the i^th observation for all points including the one deleted (the i^th point). It is computed as the square of the Pearson residual multiplied by the Hat value divided by the square of 1 minus the Hat value. Confidence interval displacement diagnostics that provides scalar measure of the influence of individual observations.
`CBAR`	`C` and `CBAR` are extensions of Cooks’ distance for logistic regression. `CBAR` measures the overall change in the fitted logits due to the deletion of the i^th observation for all points excluding the one deleted (the i^th point). It is computed as the square of the Pearson residual multiplied by the Hat value divided by (1 minus the Hat value) Confidence interval displacement diagnostic which measures the influence of deleting an individual observation.
`DIFDEV`	A statistic that measures the change in deviance that occurs when an observation is deleted from the input. It is computed as the square of the deviance residual plus `CBAR`.
`DIFCHISQ`	A statistic that measures the change in the Pearson chi-square statistic that occurs when an observation is deleted from the input. It is computed as `CBAR` divided by the Hat value.

Global Details for GLM: Linear Regression

The following table describes Global Name-Value Pairs (DM$VG) for a linear regression model.

Table 4-35 Global Details for Linear Regression

Name	Description
`ADJUSTED_R_SQUARE`	Adjusted R-Square
`AIC`	Akaike's information criterion
`COEFF_VAR`	Coefficient of variation
`CONVERGED`	Indicates whether the model build process has converged to specified tolerance. The following are the possible values: `YES` `NO`
`CORRECTED_TOTAL_DF`	Corrected total degrees of freedom
`CORRECTED_TOT_SS`	Corrected total sum of squares
`DEPENDENT_MEAN`	Dependent mean
`ERROR_DF`	Error degrees of freedom
`ERROR_MEAN_SQUARE`	Error mean square
`ERROR_SUM_SQUARES`	Error sum of squares
`F_VALUE`	Model F value statistic
`GMSEP`	Estimated mean square error of the prediction, assuming multivariate normality
`HOCKING_SP`	Hocking Sp statistic
`ITERATIONS`	Tracks the number of SGD iterations. Applicable only when the solver is SGD.
`J_P`	JP statistic (the final prediction error)
`MODEL_DF`	Model degrees of freedom
`MODEL_F_P_VALUE`	Model F value probability
`MODEL_MEAN_SQUARE`	Model mean square error
`MODEL_SUM_SQUARES`	Model sum of square errors
`NUM_PARAMS`	Number of parameters (the number of coefficients, including the intercept)
`NUM_ROWS`	Number of rows
`R_SQ`	R-Square
`RANK_DEFICIENCY`	The number of predictors excluded from the model due to multi-collinearity
`ROOT_MEAN_SQ`	Root mean square error
`SBIC`	Schwarz's Bayesian information criterion

Global Details for GLM: Logistic Regression

The following table returns Global Name-Value Pairs (DM$VG) for a logistic regression model.

Table 4-36 Global Details for Logistic Regression

Name	Description
`AIC_INTERCEPT`	Akaike's criterion for the fit of the baseline, intercept-only, model
`AIC_MODEL`	Akaike's criterion for the fit of the intercept and the covariates (predictors) mode
`CONVERGED`	Indicates whether the model build process has converged to specified tolerance. The following are the possible values: `YES` `NO`
`DEPENDENT_MEAN`	Dependent mean
`ITERATIONS`	Tracks the number of SGD iterations (number of IRLS iterations). Applicable only when the solver is SGD.
`LR_DF`	Likelihood ratio degrees of freedom
`LR_CHI_SQ`	Likelihood ratio chi-square value
`LR_CHI_SQ_P_VALUE`	Likelihood ratio chi-square probability value
`NEG2_LL_INTERCEPT`	-2 log likelihood of the baseline, intercept-only, model
`NEG2_LL_MODEL`	-2 log likelihood of the model
`NUM_PARAMS`	Number of parameters (the number of coefficients, including the intercept)
`NUM_ROWS`	Number of rows
`PCT_CORRECT`	Percent of correct predictions
`PCT_INCORRECT`	Percent of incorrectly predicted rows
`PCT_TIED`	Percent of cases where the estimated probabilities are equal for both target classes
`PSEUDO_R_SQ_CS`	Pseudo R-square Cox and Snell
`PSEUDO_R_SQ_N`	Pseudo R-square Nagelkerke
`RANK_DEFICIENCY`	The number of predictors excluded from the model due to multi-collinearity
`SC_INTERCEPT`	Schwarz's Criterion for the fit of the baseline, intercept-only, model
`SC_MODEL`	Schwarz's Criterion for the fit of the intercept and the covariates (predictors) model

Note:

When ridge regression is enabled, fewer global details are returned. For information about ridge, see Oracle Machine Learning for SQL Concepts.
When the value is NULL for a partitioned model, an exception is thrown. When the value is not null, it must contain the desired partition name.

Related Topics

Parent topic: Model Detail Views

4.9.9 Model Detail View for Multivariate State Estimation Technique - Sequential Probability Ratio Test

The model detail view specific to Multivariate State Estimation Technique - Sequential Probability Ratio Test contains information about Global Name-Value Paris.

The following are the available model views for MSET-SPRT:

Views	Description
`DM$VC`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VW`model_name	Model Build Alerts

The following table lists the Global Name-Value Pairs (DM$VGmodel_name) for an MSET-SPRT. This statistic is included when due to memory constraints MSET-SPRT cannot use the MSET_MEMORY_VECTORS value set by the user.

Table 4-37 MSET-SPRT Information in the Model Global View

Name	Description
`NUM_MVEC`	The number of memory vectors used by the model.

Parent topic: Model Detail Views

4.9.10 Model Detail Views for Naive Bayes

The model detail views specific to Naive Bayes are the prior view and result view.

These the model views available for Naive Bayes:

Model Views	Description
`DM$VB`model_name	Automatic Data Preparation Binning
`DM$VC`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VP`model_name	Naive Bayes Target Priors
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VV`model_name	Naive Bayes Conditional Probabilities
`DM$VW`model_name	Model Build Alerts

The Naive Bayes Target Priors view (DM$VPmodel_name) describes the priors of the targets for a Naive Bayes model. The view has the following columns:

Name                                      Type
----------------------------------------- ----------------------------
PARTITION_NAME                            VARCHAR2(128)
TARGET_NAME                               VARCHAR2(128)
TARGET_VALUE                              NUMBER/VARCHAR2
PRIOR_PROBABILITY                         BINARY_DOUBLE
COUNT                                     NUMBER

Table 4-38 Naive Bayes Target Priors View for Naive Bayes

Column Name	Description
`PARTITION_NAME`	The name of a feature in the model
`TARGET_NAME`	Name of the target column
`TARGET_VALUE`	Target value, numerical or categorical
`PRIOR_PROBABILITY`	Prior probability for a given `TARGET_VALUE`
`COUNT`	Number of rows for a given `TARGET_VALUE`

The Naive Bayes Conditional Probabilities view (DM$VVmodel_view) describes the conditional probabilities of the Naive Bayes model. The view has the following columns:

Name                                      Type
----------------------------------------- ----------------------------
PARTITION_NAME                            VARCHAR2(128)
TARGET_NAME                               VARCHAR2(128)
TARGET_VALUE                              NUMBER/VARCHAR2
ATTRIBUTE_NAME                            VARCHAR2(128)
ATTRIBUTE_SUBNAME                         VARCHAR2(4000)
ATTRIBUTE_VALUE                           VARCHAR2(4000)
CONDITIONAL_PROBABILITY                   BINARY_DOUBLE
COUNT                                     NUMBER

Table 4-39 Naive Bayes Conditional Probabilities View for Naive Bayes

Column Name	Description
`PARTITION_NAME`	The name of a feature in the model
`TARGET_NAME`	Name of the target column
`TARGET_VALUE`	Target value, numerical or categorical
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Machine learning attribute value for the column `ATTRIBUTE_NAME` or the nested column `ATTRIBUTE_SUBNAME` (if any).
`CONDITIONAL_PROBABILITY`	Conditional probability of a machine learning attribute for a given target
`COUNT`	Number of rows for a given machine learning attribute and a given target

The following table describes the Global Name-Value Pairs view (DM$VGmodel_name) specific to a Naive Bayes model.

Table 4-40 Global Name-Value Pairs View for Naive Bayes

Name	Description
`NUM_ROWS`	The total number of rows used in the build

Parent topic: Model Detail Views

4.9.11 Model Detail Views for Neural Network

Model detail views specific to Neural Network contain information about the weights of the neurons: input layer and hidden layers.

These are the model views available for Neural Network:

Model Views	Description
`DM$VA`model_name	Neural Network Weights
`DM$VC`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VW`model_name	Model Build Alerts

The Neural Network Weights view (DM$VAmodel_name) has the following columns:

Name                          Type                                               
----------------------        -----------------------    
PARTITION_NAME                VARCHAR2(128)         
LAYER                         NUMBER                
IDX_FROM                      NUMBER                 
ATTRIBUTE_NAME                VARCHAR2(128)         
ATTRIBUTE_SUBNAME             VARCHAR2(4000)        
ATTRIBUTE_VALUE               VARCHAR2(4000)        
IDX_TO                        NUMBER                
TARGET_VALUE                  NUMBER/VARCHAR2       
WEIGHT                        BINARY_DOUBLE

Table 4-41 Neural Network Weights View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`LAYER`	Layer ID, 0 as an input layer
`IDX_FROM`	Node index that the weight connects from (attribute id for input layer)
`ATTRIBUTE_NAME`	Attribute name (only for the input layer)
`ATTRIBUTE_SUBNAME`	Attribute subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Categorical attribute value
`IDX_TO`	Node index that the weights connects to
`TARGET_VALUE`	Target value. The value is null for regression.
`WEIGHT`	Value of the weight

The view Global Name-Value Pairs (DM$VGmodel_name) is a pre-existing view. The following name-value pairs are specific to a Neural Network view.

Table 4-42 Global Name-Value Pairs Viewfor Neural Network

Name	Description
`CONVERGED`	Indicates whether the model build process has converged to specified tolerance. The following are the possible values: `YES` `NO`
`ITERATIONS`	Number of iterations
`LOSS_VALUE`	Loss function value (if it is with `NNET_REGULARIZER_HELDASIDE` regularization, it is the loss function value on test data)
`NUM_ROWS`	Number of rows in the model (or partitioned model)

Parent topic: Model Detail Views

4.9.12 Model Detail Views for Random Forest

Model detail views specific to Random Forest contain variable importance measures and statistics.

The following model detail views are available for Random Forest:

Model View	Description
`DM$VA`model_name	Variable Importance
`DM$VC`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VW`model_name	Model Build Alerts

Model detail views and statistics specific to Random Forest are:

Variable Importance statistics DM$VAmodel_name
Random Forest statistics in the Global Name-Value Pairs DM$VGmodel_name view

One of the important outputs from a Random Forest model build is a ranking of attributes based on their relative importance. This is measured using Mean Decrease Gini. The DM$VAmodel_name view has the following columns:


Name                              Type        
------------------------          --------------- 
PARTITION_NAME                    VARCHAR2(128)   
ATTRIBUTE_NAME                    VARCHAR2(128)   
ATTRIBUTE_SUBNAME                 VARCHAR2(128)   
ATTRIBUTE_IMPORTANCE              BINARY_DOUBLE

Table 4-43 Variable Importance Model View

Column Name	Description
`PARTITION_NAME`	Partition name. The value is null for models which are not partitioned.
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_IMPORTANCE`	Measure of importance for an attribute in the forest (mean Decrease Gini value)

The Global Name-Value Pairs (DM$VGmodel_name) view is a pre-existing view. The following name-value pairs are added to the view.

Table 4-44 Random Forest Statistics Information In Model Global View

Name	Description
`AVG_DEPTH`	Average depth of the trees in the forest
`AVG_NODECOUNT`	Average number of nodes per tree
`MAX_DEPTH`	Maximum depth of the trees in the forest
`MAX_NODECOUNT`	Maximum number of nodes per tree
`MIN_DEPTH`	Minimum depth of the trees in the forest
`MIN_NODECOUNT`	Minimum number of nodes per tree
`NUM_ROWS`	The total number of rows used in the build

Parent topic: Model Detail Views

4.9.13 Model Detail View for Support Vector Machine

Model detail views specific to Support Vector Machine (SVM) contain linear coefficients and support vector statistics.

These model views are available for SVM:

Model Views	Description
`DM$VCS`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VW`model_name	Model Build Alerts

The linear coefficient view DM$VLmodel_name describes the coefficients of a linear SVM algorithm. The target_value field in the view is present only for classification and has the type of the target. Regression models do not have a target_value field.

The reversed_coefficient field shows the value of the coefficient after reversing the automatic data preparation transformations. If data preparation is disabled, then coefficient and reversed_coefficient have the same value. The view has the following columns:

Name                                       Type
----------------------------------------- --------------------------------
PARTITION_NAME                            VARCHAR2(128)
TARGET_VALUE                              NUMBER/VARCHAR2
ATTRIBUTE_NAME                            VARCHAR2(128)
ATTRIBUTE_SUBNAME                         VARCHAR2(4000)
ATTRIBUTE_VALUE                           VARCHAR2(4000)
COEFFICIENT                               BINARY_DOUBLE
REVERSED_COEFFICIENT                      BINARY_DOUBLE

Table 4-45 Linear Coefficient View for Support Vector Machine

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`TARGET_VALUE`	Target value, numerical or categorical
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Value of a categorical attribute
`COEFFICIENT`	Projection coefficient value
`REVERSED_COEFFICIENT`	Coefficient transformed on the original scale

The following table describes the SVM statistics global view.

Table 4-46 Support Vector Statistics Information In Model Global View

Name	Description
`CONVERGED`	Indicates whether the model build process has converged to specified tolerance: `YES` `NO`
`ITERATIONS`	Number of iterations performed during build
`NUM_ROWS`	Number of rows used for the build
`REMOVED_ROWS_ZERO_NORM`	Number of rows removed due to 0 norm. This applies to one-class linear models only.

Parent topic: Model Detail Views

4.9.14 Model Detail Views for XGBoost

The model detail views specific to XGBoost contain information about Feature Importance view and Global Name-Value Pairs view.

The following are the available model views for XGBoost Classification:

Model Views	Description
`DM$VC`model_name	Scoring Cost Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VI`model_name	XGBoost Attribute Importance
`DM$VS`model_name	Computed Settings
`DM$VT`model_name	Classification Targets
`DM$VW`model_name	Model Build Alerts

The following are the available model views for XGBoost Regression:

Views	Description
`DM$VG`model_name	Global Name-Value Pairs
`DM$VI`model_name	XGBoost Attribute Importance
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts

The DM$VImodel_name view reports the feature importance values for each attribute of each partition of the model.

The view has the following columns for tree models (gbtree and dart boosters).

Name               Type
-----------------  --------------
PNAME              VARCHAR2(128)
ATTRIBUTE_NAME     VARCHAR2(128)
ATTRIBUTE_SUBNAME  VARCHAR2(4000)
ATTRIBUTE_VALUE    VARCHAR2(4000)
GAIN               BINARY_DOUBLE
COVER              BINARY_DOUBLE
FREQUENCY          BINARY_DOUBLE

Table 4-47 Feature Importance View for a Tree Model

Column Name	Description
`PNAME`	The name of a partition in a partitioned model.
`ATTRIBUTE_NAME`	The column name.
`ATTRIBUTE_SUBNAME`	The nested column subname; the value is null for non-nested columns.
`ATTRIBUTE_VALUE`	The value of a categorical attribute.
`GAIN`	The fractional contribution of each feature to the model based on the total gain of a feature’s splits; a higher percentage means a more important predictive feature.
`COVER`	The number of observation either seen by a split or collected by a leaf during training.
`FREQUENCY`	A percentage representing the relative number of times a feature has been used in trees.

For a linear model (gblinear) booster, the feature importance is the absolute magnitude of linear coefficients.

The view has the following columns for linear models.

Name        Type
-----------------  --------------
PNAME              VARCHAR2(128)
ATTRIBUTE_NAME     VARCHAR2(128)
ATTRIBUTE_SUBNAME  VARCHAR2(4000)
ATTRIBUTE_VALUE    VARCHAR2(4000)
WEIGHT             BINARY_DOUBLE
CLASS              BINARY_DOUBLE

Table 4-48 Feature Importance View for a Linear Model

Column Name	Description
`PNAME`	The name of a partition in a partitioned model.
`ATTRIBUTE_NAME`	The column name.
`ATTRIBUTE_SUBNAME`	The nested column subname; the value is null for non-nested columns.
`ATTRIBUTE_VALUE`	The value of a categorical attribute.
`WEIGHT`	The linear coefficient of the feature.
`CLASS`	The class label for a multiclass model.

The DM$VGmodel_name view reports global statistics for an XGBoost model. The statistics include an evaluation of the training data set using the evaluation metric you specified with the learning task eval_metric setting, or the default eval_metric if you didn't specify one. The view displays only the result of the last training iteration. When you specify more than one eval_metric, the view contains multiple rows, one for each eval_metric.

Parent topic: Model Detail Views

4.9.15 Model Detail Views for Clustering Algorithms

Oracle Machine Learning for SQL supports these clustering algorithms: Expectation Maximization (EM), k-Means (KM), and orthogonal partitioning clustering (O-Cluster, OC).

All clustering algorithms share the following views:

Model Views	Description
`DM$VD`model_name:	Clustering Description
`DM$VA`model_name	Clustering Attribute Statistics
`DM$VH`model_name	Clustering Histograms
`DM$VR`model_name	Clustering Rules

The Cluster Description view DM$VDmodel_name describes cluster level information about a clustering model. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
CLUSTER_ID                         NUMBER
CLUSTER_NAME                       NUMBER/VARCHAR2  
RECORD_COUNT                       NUMBER
PARENT                             NUMBER           
TREE_LEVEL                         NUMBER           
LEFT_CHILD_ID                      NUMBER           
RIGHT_CHILD_ID                     NUMBER

Table 4-49 Clustering Description View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`CLUSTER_ID`	The ID of a cluster in the model
`CLUSTER_NAME`	Specifies the label of the cluster
`RECORD_COUNT`	Specifies the number of records
`PARENT`	The ID of the parent
`TREE_LEVEL`	Specifies the number of splits from the root
`LEFT_CHILD_ID`	The ID of the child cluster on the left side of the split
`RIGHT_CHILD_ID`	The ID of the child cluster on the right side of the split

The attribute view DM$VAmodel_name describes attribute level information about a clustering model. The values of the mean, variance, and mode for a particular cluster can be obtained from this view. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
CLUSTER_ID                         NUMBER
CLUSTER_NAME                       NUMBER/VARCHAR2
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
MEAN                               BINARY_DOUBLE
VARIANCE                           BINARY_DOUBLE
MODE_VALUE                         VARCHAR2(4000)

Table 4-50 Clustering Attribute Statistics

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model
`CLUSTER_ID`	The ID of a cluster in the model
`CLUSTER_NAME`	Specifies the label of the cluster
`ATTRIBUTE_NAME`	Specifies the attribute name
`ATTRIBUTE_SUBNAME`	Specifies the attribute subname. For vector data types, this attribute shows each vector dimension as an individual predictor with `DM$$xxx`, where xxx is the position of the vector.
`MEAN`	The field returns the average value of a numeric attribute
`VARIANCE`	The variance of a numeric attribute
`MODE_VALUE`	The mode is the most frequent value of a categorical attribute

The histogram view DM$VHmodel_name describes histogram level information about a clustering model. The bin information as well as bin counts can be obtained from this view. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
CLUSTER_ID                         NUMBER
CLUSTER_NAME                       NUMBER/VARCHAR2
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
BIN_ID                             NUMBER
LOWER_BIN_BOUNDARY                 BINARY_DOUBLE                
UPPER_BIN_BOUNDARY                 BINARY_DOUBLE         
ATTRIBUTE_VALUE                    VARCHAR2(4000)       
COUNT                              NUMBER

Table 4-51 Clustering Histograms View

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model
`CLUSTER_ID`	The ID of a cluster in the model
`CLUSTER_NAME`	Specifies the label of the cluster
`ATTRIBUTE_NAME`	Specifies the attribute name
`ATTRIBUTE_SUBNAME`	Specifies the attribute subname
`BIN_ID`	Bin ID
`LOWER_BIN_BOUNDARY`	Numeric lower bin boundary
`UPPER_BIN_BOUNDARY`	Numeric upper bin boundary
`ATTRIBUTE_VALUE`	Categorical attribute value
`COUNT`	Histogram count

The rule view DM$VRmodel_name describes the rule level information about a clustering model. The information is provided at attribute predicate level. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
CLUSTER_ID                         NUMBER              
CLUSTER_NAME                       NUMBER/VARCHAR2
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
OPERATOR                           VARCHAR2(2)         
NUMERIC_VALUE                      NUMBER              
ATTRIBUTE_VALUE                    VARCHAR2(4000)      
SUPPORT                            NUMBER              
CONFIDENCE                         BINARY_DOUBLE       
RULE_SUPPORT                       NUMBER              
RULE_CONFIDENCE                    BINARY_DOUBLE

Table 4-52 Clustering Rules View

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model
`CLUSTER_ID`	The ID of a cluster in the model
`CLUSTER_NAME`	Specifies the label of the cluster
`ATTRIBUTE_NAME`	Specifies the attribute name
`ATTRIBUTE_SUBNAME`	Specifies the attribute subname
`OPERATOR`	Attribute predicate operator - a conditional operator taking the following values: IN, = , <>, < , >, <=, and >=
`NUMERIC_VALUE`	Numeric lower bin boundary
`ATTRIBUTE_VALUE`	Categorical attribute value
`SUPPORT`	Attribute predicate support
`CONFIDENCE`	Attribute predicate confidence
`RULE_SUPPORT`	Rule level support
`RULE_CONFIDENCE`	Rule level confidence

Parent topic: Model Detail Views

4.9.16 Model Detail Views for Expectation Maximization

Model detail views specific to Expectation Maximization (EM) contain additional information about an EM model. Additional views are available for EM Clustering, but are absent for EM Anomaly.

These are the model views available for Expectation Maximization:

Model Views	Description
`DM$VA`model_name	Clustering Attribute Statistics
`DM$VB`model_name	Attribute Pair Kullback-Leibler Divergence
`DM$VD`model_name	Clustering Description
`DM$VF`model_name	Expectation Maximization Bernoulli parameters
`DM$VG`model_name	Global Name-Value Pairs
`DM$VH`model_name	Clustering Histograms
`DM$VI`model_name	Unsupervised Attribute Importance
`DM$VM`model_name	Expectation Maximization Gaussian parameters
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VO`model_name	Expectation Maximization Components
`DM$VP`model_name	Expectation Maximization Projections
`DM$VR`model_name	Clustering Rules
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts

For EM Clustering model, the following views contain information that is not in the clustering views. For the clustering views, refer to "Model Detail Views for Clustering Algorithms".

The Expectation Maximization Components view (DM$VOmodel_name) describes the EM Cluster components. The component view contains information about their prior probabilities and what cluster they map to. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
COMPONENT_ID                       NUMBER
CLUSTER_ID                         NUMBER
PRIOR_PROBABILITY                  BINARY_DOUBLE

Table 4-53 Expectation Maximization Components View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`COMPONENT_ID`	Unique identifier of a component
`CLUSTER_ID`	The ID of a cluster in the model
`PRIOR_PROBABILITY`	Component prior probability

The Expectation Maximization Gaussian view (DM$VMmodel_name) provides information about the mean and variance parameters for the attributes by Gaussian distribution models. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
COMPONENT_ID                       NUMBER
ATTRIBUTE_NAME                     VARCHAR2(4000)
MEAN                               BINARY_DOUBLE
VARIANCE                           BINARY_DOUBLE

The Expectation Maximization Bernoulli parameters view (DM$VFmodel_name) provides information about the parameters of the multi-valued Bernoulli distributions used by the EM model. The view has the following columns:

Name                                Type
 ---------------------------------- ----------------------------
 PARTITION_NAME                     VARCHAR2(128)
 COMPONENT_ID                       NUMBER
 ATTRIBUTE_NAME                     VARCHAR2(4000)
 ATTRIBUTE_VALUE                    VARCHAR2(4000)
 FREQUENCY                          BINARY_DOUBLE

Table 4-54 Expectation Maximization Bernoulli parameters View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`COMPONENT_ID`	Unique identifier of a component
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_VALUE`	Categorical attribute value
`FREQUENCY`	The frequency of the multivalued Bernoulli distribution for the attribute/value combination specified by `ATTRIBUTE_NAME` and `ATTRIBUTE_VALUE`.

For 2-Dimensional columns, EM provides an attribute ranking similar to that of attribute importance. This ranking is based on a rank-weighted average over Kullback–Leibler divergence computed for pairs of columns. This unsupervised attribute importance is shown in the Unsupervised Attribute Importance view (DM$VImodel_name) and has the following columns:

Name                                      Type
----------------------------------------- ----------------------------
PARTITION_NAME                            VARCHAR2(128)
ATTRIBUTE_NAME                            VARCHAR2(128)
ATTRIBUTE_IMPORTANCE_VALUE                BINARY_DOUBLE
ATTRIBUTE_RANK                            NUMBER

Table 4-55 Unsupervised Attribute Importance View for Expectation Maximization

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_IMPORTANCE_VALUE`	Importance value
`ATTRIBUTE_RANK`	An attribute rank based on the importance value

The pairwise Kullback–Leibler divergence is reported in the Attribute Pair Kullback-Leibler Divergence view (DM$VBmodel_name). This metric evaluates how much the observed joint distribution of two attributes diverges from the expected distribution under the assumption of independence. That is, the higher the value, the more dependent the two attributes are. The dependency value is scaled based on the size of the grid used for each pairwise computation. That ensures that all values fall within the [0; 1] range and are comparable. The view has the following columns:

Name                                      Type
----------------------------------------- ----------------------------
PARTITION_NAME                            VARCHAR2(128)
ATTRIBUTE_NAME_1                          VARCHAR2(128)
ATTRIBUTE_NAME_2                          VARCHAR2(128)
DEPENDENCY                                BINARY_DOUBLE

Table 4-56 Attribute Pair Kullback-Leibler Divergence View for Expectation Maximization

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ATTRIBUTE_NAME_1`	Name of the first attribute
`ATTRIBUTE_NAME_2`	Name of the second attribute
`DEPENDENCY`	Scaled pairwise Kullback-Leibler divergence

The projection table DM$VPmodel_name shows the coefficients used by random projections to map nested columns to a lower dimensional space. The view has rows only when nested or text data is present in the build data. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
FEATURE_NAME                       VARCHAR2(4000)
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
ATTRIBUTE_VALUE                    VARCHAR2(4000)
COEFFICIENT                        NUMBER

Table 4-57 Projection table for Expectation Maximization

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`FEATURE_NAME`	Name of feature
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Categorical attribute value
`COEFFICIENT`	Projection coefficient. The representation is sparse; only the non-zero coefficients are returned.

For EM Anomaly, currently there are no additional views other than the classification views. For the classification view, refer to “Model Detail Views for Classification Algorithms”.

Global Details for Expectation Maximization

The following table describes global details for EM.

Table 4-58 Global Details for Expectation Maximization

Name	Description
`CONVERGED`	Indicates whether the model build process has converged to specified tolerance. The possible values are: `YES` `NO`
`LOGLIKELIHOOD`	Loglikelihood on the build data
`NUM_COMPONENTS`	Number of components produced by the model
`NUM_CLUSTERS`	Number of clusters produced by the model (only available for EM Clustering)
`NUM_ROWS`	Number of rows used in the build
`RANDOM_SEED`	The random seed value used for the model build
`REMOVED_COMPONENTS`	The number of empty components excluded from the model

Related Topics

Model Detail Views for Clustering Algorithms

Parent topic: Model Detail Views

4.9.17 Model Detail Views for k-Means

Model detail views specific to k-Means (KM) contain clustering description view (DM$VG), and scoring information.

The following model views are available for k-Means algorithm.

Model Views	Description
`DM$VA`model_name	Clustering Attribute Statistics
`DM$VC`model_name	k-Means Scoring Centroids
`DM$VD`model_name	Clustering Description
`DM$VG`model_name	Global Name-Value Pairs
`DM$VH`model_name	Clustering Histograms
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VR`model_name	Clustering Rules
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts

"Model Detail Views for Clustering Algorithms" discusses common model views across clustering algorithms. Global Name-Value Pairs view (DM$VG), which contains information about Computed Settings view (DM$VS) and Model Build Alerts view (DM$VW), and Normalization and Missing Value Handling view (DM$VN) are addressed individually.

The following views contain information that is specific to k-Means model.

The k-Means Clustering Description view DM$VDmodel_name has an additional column:

Name                               Type
---------------------------------- ----------------------------
DISPERSION                         BINARY_DOUBLE

Table 4-59 Clustering Description for k-Means

Column Name	Description
`DISPERSION`	A measure used to quantify whether a set of observed occurrences are dispersed compared to a standard statistical model.

The k-Means Scoring Centroids view DM$VCmodel_name describes the centroid of each leaf clusters:

Name                                Type
 ---------------------------------- ----------------------------
 PARTITION_NAME                     VARCHAR2(128)
 CLUSTER_ID                         NUMBER
 CLUSTER_NAME                       NUMBER/VARCHAR2
 ATTRIBUTE_NAME                     VARCHAR2(128)
 ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
 ATTRIBUTE_VALUE                    VARCHAR2(4000)
 VALUE                              BINARY_DOUBLE

Table 4-60 k-Means Scoring Centroids View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`CLUSTER_ID`	The ID of a cluster in the model
`CLUSTER_NAME`	Specifies the label of the cluster
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Categorical attribute value
`VALUE`	Specifies the centroid value

The following table describes Global Name-Value Pairs view (DM$VG) for k-Means.

Table 4-61 k–Means Global Name-Value Pairs View

Name	Description
`CONVERGED`	Indicates whether the model build process has converged to specified tolerance. The following are the possible values: `YES` `NO`
`NUM_ROWS`	Number of rows used in the build
`REMOVED_ROWS_ZERO_NORM`	Number of rows removed due to 0 norm. This applies only to models using cosine distance.

Related Topics

Parent topic: Model Detail Views

4.9.18 Model Detail Views for O-Cluster

Model detail views specific to O-Cluster (OC) contain information about description view, histograms view, and global view.

These are the available model views for O-Cluster:

Model Views	Description
`DM$VA`model_name	Clustering Attribute Statistics
`DM$VB`model_name	Automatic Data Preparation Binning
`DM$VD`model_name	Clustering Description
`DM$VG`model_name	Global Name-Value Pairs
`DM$VH`model_name	Clustering Histograms
`DM$VR`model_name	Clustering Rules
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts

The following views contain information that is specific to an O-Cluster model. For the clustering views, refer to "Model Detail Views for Clustering Algorithms". The OC algorithm uses the same descriptive statistics views as Expectation Maximization (EM) and k-Means (KM). The following are the statistics views:

The Cluster Description view (DM$VDmodel_name) describes the O-Cluster components. The Cluster Description view has additional fields that specify the split predicate. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
OPERATOR                           VARCHAR2(2)        
VALUE                              SYS.XMLTYPE

Table 4-62 Cluster Description View for O-Cluster

Column Name	Description
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`OPERATOR`	Split operator
`VALUE`	List of split values

The structure of the SYS.XMLTYPE is as follows:

<Element>splitval1</Element>

The OC algorithm uses a Clustering Histograms view (DM$VHmodel_name) with different columns than EM and KM. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITON_NAME                      VARCHAR2(128)
CLUSTER_ID                         NUMBER
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
BIN_ID                             NUMBER
LABEL                              VARCHAR2(4000)   
COUNT                              NUMBER

Table 4-63 Clustering Histograms View for O-Cluster

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`CLUSTER_ID`	Unique identifier of a component
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`BIN_ID`	Unique identifier
`LABEL`	Bin label
`COUNT`	Bin histogram count

The following table describes the Global Name-Value Pairs (DM$VGmodel_name) view specific to O-Cluster.

Table 4-64 O-Cluster Statistics Information In Model Global View

Name	Description
`NUM_ROWS`	The total number of rows used in the build

Related Topics

Model Detail Views for Clustering Algorithms

Parent topic: Model Detail Views

4.9.19 Model Detail Views for Explicit Semantic Analysis

Model detail views specific to Explicit Semantic Analysis (ESA) contain information about attribute statistics and features.

These are the available model views:

Model Views	Description
`DM$VA`model_name	Explicit Semantic Analysis Matrix
`DM$VF`model_name	Explicit Semantic Analysis Features
`DM$VG`model_name	Global Name-Value Pairs
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts
`DM$VX`model_name	Text Features

Explicit Semantic Analysis Matrix (DM$VAmodel_name): This view has different columns for feature extraction and classification. For feature extraction, this view contains model attribute coefficients per feature. For classification, this view contains model attribute coefficients per target class.
Explicit Semantic Analysis Features (DM$VFmodel_name): This view is applicable only for feature extraction.

The Explicit Semantic Analysis Matrix view (DM$VAmodel_name) has the following columns for feature extraction:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
FEATURE_ID                         NUMBER/VARHCAR2, DATE, TIMESTAMP, 
                                   TIMESTAMP WITH TIME ZONE,
                                   TIMESTAMP WITH LOCAL TIME ZONE 
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
ATTRIBUTE_VALUE                    VARCHAR2(4000)
COEFFICIENT                        BINARY_DOUBLE

Table 4-65 Explicit Semantic Analysis Matrix for Feature Extraction

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`FEATURE_ID`	Unique identifier of a feature as it appears in the training data
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Categorical attribute value
`COEFFICIENT`	A measure of the weight of the attribute with respect to the feature

The (DM$VAmodel_name) view comprises of attribute coefficients for all target classes.

The view Explicit Semantic Analysis Matrix (DM$VAmodel_name) has the following columns for classification:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
TARGET_VALUE                       NUMBER/VARCHAR2
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
ATTRIBUTE_VALUE                    VARCHAR2(4000)
COEFFICIENT                        BINARY_DOUBLE

Table 4-66 Explicit Semantic Analysis Matrix for Classification

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`TARGET_VALUE`	Value of the target
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Categorical attribute value
`COEFFICIENT`	A measure of the weight of the attribute with respect to the feature

The Explicit Semantic Analysis Features view (DM$VFmodel_name) has a unique row for every feature in one view. This feature is helpful if the model was pre-built and the source training data are not available. The view has the following columns:


Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
FEATURE_ID                         NUMBER/VARHCAR2, DATE, TIMESTAMP, 
                                   TIMESTAMP WITH TIME ZONE,
                                   TIMESTAMP WITH LOCAL TIME ZONE

Table 4-67 Explicit Semantic Analysis Features for Explicit Semantic Analysis

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`FEATURE_ID`	Unique identifier of a feature as it appears in the training data

The following table describes the Global Name-Value Pairs view (DM$VGmodel_name) specific to ESA.

Table 4-68 Explicit Semantic Analysis Statistics Information In Model Global View

Name	Description
`NUM_ROWS`	The total number of input rows
`REMOVED_ROWS_BY_FILTERS`	Number of rows removed by filters

Parent topic: Model Detail Views

4.9.20 Model Detail Views for Non-Negative Matrix Factorization

Model detail views specific to Non-Negative Matrix Factorization (NMF) contain information about the encoding H matrix and H inverse matrix.

These are the available model views for NMF:

Model Views	Description
`DM$VE`model_name	Non-Negative Matrix Factorization H Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VI`model_name	Non-Negative Matrix Factorization Inverse H Matrix
`DM$VN`model_name	Normalization andMissing Value Handling
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts

The views specific to NMF are:

Non-Negative Matrix Factorization H Matrix view (DM$VEmodel_name)
Non-Negative Matrix Factorization Inverse H Matrix view (DM$VImodel_name)

The view DM$VEmodel_name describes the encoding (H) matrix of an NMF model. The FEATURE_NAME column type may be either NUMBER or VARCHAR2. The view has the following columns.

Name                 Type
-------------------  --------------------------
PARTITION_NAME       VARCHAR2(128)
FEATURE_ID           NUMBER
FEATURE_NAME         NUMBER/VARCHAR2
ATTRIBUTE_NAME       VARCHAR2(128)
ATTRIBUTE_SUBNAME    VARCHAR2(4000)
ATTRIBUTE_VALUE      VARCHAR2(4000)
COEFFICIENT          BINARY_DOUBLE

Table 4-69 Non-Negative Matrix Factorization H Matrix View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`FEATURE_ID`	The ID of a feature in the model
`FEATURE_NAME`	The name of a feature in the model
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Specifies the value of attribute
`COEFFICIENT`	The attribute encoding that represents its contribution to the feature

The view DM$VImodel_view describes the inverse H matrix of an NMF model. The FEATURE_NAME column type may be either NUMBER or VARCHAR2. The view has the following schema:

Name               Type
-----------------  ------------------------
PARTITION_NAME     VARCHAR2(128)
FEATURE_ID         NUMBER
FEATURE_NAME       NUMBER/VARCHAR2
ATTRIBUTE_NAME     VARCHAR2(128)
ATTRIBUTE_SUBNAME  VARCHAR2(4000)
ATTRIBUTE_VALUE    VARCHAR2(4000)
COEFFICIENT        BINARY_DOUBLE

Table 4-70 Non-Negative Matrix Factorization Inverse H Matrix View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`FEATURE_ID`	The ID of a feature in the model
`FEATURE_NAME`	The name of a feature in the model
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Specifies the value of attribute
`COEFFICIENT`	The attribute encoding that represents its contribution to the feature

The following table describes the Global Name-Value Pairs view (DM$VGmodel_name) specific to NMF.

Table 4-71 Global Name-Value Pairs View for NMF

Name	Description
`CONV_ERROR`	Convergence error
`CONVERGED`	Indicates whether the model build process has converged to specified tolerance. The following are the possible values: `YES` `NO`
`ITERATIONS`	Number of iterations performed during build
`NUM_ROWS`	Number of rows used in the build input data set
`SAMPLE_SIZE`	Number of rows used by the build

Parent topic: Model Detail Views

4.9.21 Model Detail Views for Singular Value Decomposition

Model detail views specific to Singular Value Decomposition (SVD) contain information about the S matrix, right-singular vectors, and left-singular vectors.

These are the available model views for SVD:

Model Views	Description
`DM$VE`model_name	Singular Value Decomposition S Matrix
`DM$VG`model_name	Global Name-Value Pairs
`DM$VN`model_name	Normalization and Missing Value Handling
`DM$VS`model_name	Computed Settings
`DM$VU`model_name	Singular Value Decomposition U Matrix
`DM$VV`model_name	Singular Value Decomposition V Matrix
`DM$VW`model_name	Model Build Alerts

The Singular Value Decomposition S Matrix view (DM$VEmodel_name) leverages the fact that each singular value in the SVD model has a corresponding principal component in the associated Principal Components Analysis (PCA) model to relate a common set of information for both classes of models. For an SVD model, it describes the content of the S matrix. When PCA scoring is selected as a build setting, the variance and percentage cumulative variance for the corresponding principal components are shown as well. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
FEATURE_ID                         NUMBER
FEATURE_NAME                       NUMBER/VARCHAR2
VALUE                              BINARY_DOUBLE
VARIANCE                           BINARY_DOUBLE  
PCT_CUM_VARIANCE                   BINARY_DOUBLE

Table 4-72 Singular Value Decomposition S Matrix View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`FEATURE_ID`	The ID of a feature in the model
`FEATURE_NAME`	The name of a feature in the model
`VALUE`	The matrix entry value
`VARIANCE`	The variance explained by a component. This column is only present for SVD models with setting `dbms_data_mining.svds_scoring_mode` set to `dbms_data_mining.svds_scoring_pca` This column is non-null only if the build data is centered, either manually or because of the following setting:`dbms_data_mining.prep_auto` is set to `dbms_data_mining.prep_auto_on`.
`PCT_CUM_VARIANCE`	The percent cumulative variance explained by the components thus far. The components are ranked by the explained variance in descending order. This column is only present for SVD models with setting `dbms_data_mining.svds_scoring_mode` set to `dbms_data_mining.svds_scoring_pca` This column is non-null only if the build data is centered, either manually or because of the following setting:`dbms_data_mining.prep_auto` is set to `dbms_data_mining.prep_auto_on`.

The Singular Value Decomposition V Matrix view (DM$VVmodel_view) describes the right-singular vectors of an SVD model. For a PCA model it describes the principal components (eigenvectors). The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
FEATURE_ID                         NUMBER
FEATURE_NAME                       NUMBER/VARCHAR2
ATTRIBUTE_NAME                     VARCHAR2(128)
ATTRIBUTE_SUBNAME                  VARCHAR2(4000)
ATTRIBUTE_VALUE                    VARCHAR2(4000)
VALUE                              BINARY_DOUBLE

Table 4-73 Singular Value Decomposition V Matrix View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`FEATURE_ID`	The ID of a feature in the model
`FEATURE_NAME`	The name of a feature in the model
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_VALUE`	Categorical attribute value. For numerical attributes, `ATTRIBUTE_VALUE` is null.
`VALUE`	The matrix entry value

The Singular Value Decomposition U Matrix view (DM$VUmodel_name) describes the left-singular vectors of an SVD model. For a PCA model, it describes the projection of the data in the principal components. This view does not exist unless the settings dbms_data_mining.svds_u_matrix_output is set to dbms_data_mining.svds_u_matrix_enable. The view has the following columns:

Name                               Type
---------------------------------- ----------------------------
PARTITION_NAME                     VARCHAR2(128)
CASE_ID                            NUMBER/VARHCAR2, DATE, TIMESTAMP, 
                                   TIMESTAMP WITH TIME ZONE,
                                   TIMESTAMP WITH LOCAL TIME ZONE
FEATURE_ID                         NUMBER
FEATURE_NAME                       NUMBER/VARCHAR2
VALUE                              BINARY_DOUBLE

Table 4-74 Singular Value Decomposition U Matrix View or Projection Data in Principal Components

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`CASE_ID`	Unique identifier of the row in the build data described by the U matrix projection.
`FEATURE_ID`	The ID of a feature in the model
`FEATURE_NAME`	The name of a feature in the model
`VALUE`	The matrix entry value

Global Details for Singular Value Decomposition

The following table describes the Global Name-Value Pairs view (DM$VGmodel_name) specific to a SVD model.

Table 4-75 Global Name-Value Pairs View for Singular Value Decomposition

Name	Description
`NUM_COMPONENTS`	Number of features (components) produced by the model
`NUM_ROWS`	The total number of rows used in the build
`SUGGESTED_CUTOFF`	Suggested cutoff that indicates how many of the top computed features capture most of the variance in the model. Using only the features below this cutoff would be a reasonable strategy for dimensionality reduction.

Related Topics

Oracle Database PL/SQL Packages and Types Reference

Parent topic: Model Detail Views

4.9.22 Model Detail Views for Minimum Description Length

Model detail views specific to Minimum Description Length (MDL) (for calculating attribute importance) contain information about attribute importance models.

These are the available model views for MDL:

Model Views	Description
`DM$VA`model_name	Attribute Importance
`DM$VB`model_name	Automatic Data Preparation Binning
`DM$VG`model_name	Global Name-Value Pairs
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts

The Attribute Importance view (DM$VAmodel_name) describes the attribute importance as well as the attribute importance rank. The view has the following columns:

Name                                      Type
----------------------------------------- ----------------------------
PARTITION_NAME                            VARCHAR2(128)
ATTRIBUTE_NAME                            VARCHAR2(128)
ATTRIBUTE_SUBNAME                         VARCHAR2(4000)
ATTRIBUTE_IMPORTANCE_VALUE                BINARY_DOUBLE
ATTRIBUTE_RANK                            NUMBER

Table 4-76 Attribute Importance View for Minimum Description Length

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`ATTRIBUTE_IMPORTANCE_VALUE`	Importance value
`ATTRIBUTE_RANK`	Rank based on importance

The following table describes the Global Name-Value Pairs view (DM$VGmodel_name) specific to MDL.

Table 4-77 Global Name-Value Pairs View for MDL

Name	Description
`NUM_ROWS`	The total number of rows used in the build

Parent topic: Model Detail Views

4.9.23 Model Detail Views for Binning

The binning view DM$VB describes the bin boundaries used in automatic data preparation.

The view has the following columns:

Name                       Type
--------------------       --------------------
PARTITION_NAME             VARCHAR2(128)
ATTRIBUTE_NAME             VARCHAR2(128)   
ATTRIBUTE_SUBNAME          VARCHAR2(4000)  
BIN_ID                     NUMBER           
LOWER_BIN_BOUNDARY         BINARY_DOUBLE          
UPPER_BIN_BOUNDARY         BINARY_DOUBLE   
ATTRIBUTE_VALUE            VARCHAR2(4000)

Table 4-78 Model Details View for Binning

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ATTRIBUTE_NAME`	Specifies the attribute name
`ATTRIBUTE_SUBNAME`	Specifies the attribute subname
`BIN_ID`	Bin ID (or bin identifier)
`LOWER_BIN_BOUNDARY`	Numeric lower bin boundary
`UPPER_BIN_BOUNDARY`	Numeric upper bin boundary
`ATTRIBUTE_VALUE`	Categorical value

Parent topic: Model Detail Views

4.9.24 Model Detail Views for Global Information

Model detail views for global information contain information about global statistics, alerts, and computed settings.

The Global Name-Value Pairs view (DM$VGmodel_name) describes global statistics related to the model build. Examples include the number of rows used in the build, the convergence status, and the model quality metrics. The view has the following columns:

Name                        Type
-------------------         --------------------
PARTITION_NAME              VARCHAR2(128)
NAME                        VARCHAR2(30)     
NUMERIC_VALUE               NUMBER           
STRING_VALUE                VARCHAR2(4000)

Table 4-79 Global Name-Value Pairs View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`NAME`	Name of the statistic
`NUMERIC_VALUE`	Numeric value of the statistic
`STRING_VALUE`	Categorical value of the statistic

The Model Build Alerts view (DM$VWmodel_name) lists alerts issued during the model build. The view has the following columns:

Name                      Type
-------------------       ----------------------
PARTITION_NAME            VARCHAR2(128)
ERROR_NUMBER              BINARY_DOUBLE
ERROR_TEXT                VARCHAR2(4000)

Table 4-80 Model Build Alerts View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`ERROR_NUMBER`	Error number (valid when event is Error)
`ERROR_TEXT`	Error message

The Computed Settings view (DM$VSmodel_name) lists the algorithm computed settings. The view has the following columns:

Name                       Type
-----------------          --------------------
PARTITION_NAME             VARCHAR2(128)
SETTING_NAME               VARCHAR2(30)
SETTING_VALUE              VARCHAR2(4000)

Table 4-81 Computed Settings View

Column Name	Description
`PARTITION_NAME`	Partition name in a partitioned model
`SETTING_NAME`	Name of the setting
`SETTING_VALUE`	Value of the setting

Parent topic: Model Detail Views

4.9.25 Model Detail Views for Normalization and Missing Value Handling

The Normalization and Missing Value Handling view DM$VN describes the normalization parameters used in Automatic Data Preparation (ADP) and the missing value replacement when a NULL value is encountered. Missing value replacement applies only to the twodimensional columns and does not apply to the nested columns.

The view has the following columns:

Name                         Type
----------------------       -----------------------
PARTITION_NAME               VARCHAR2(128)
ATTRIBUTE_NAME               VARCHAR2(128)
ATTRIBUTE_SUBNAME            VARCHAR2(4000)
NUMERIC_MISSING_VALUE        BINARY_DOUBLE  
CATEGORICAL_MISSING_VALUE    VARCHAR2(4000)  
NORMALIZATION_SHIFT          BINARY_DOUBLE   
NORMALIZATION_SCALE          BINARY_DOUBLE

Table 4-82 Normalization and Missing Value Handling View

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model
`ATTRIBUTE_NAME`	Column name
`ATTRIBUTE_SUBNAME`	Nested column subname. The value is null for non-nested columns.
`NUMERIC_MISSING_VALUE`	Numeric missing value replacement
`CATEGORICAL_MISSING_VALUE`	Categorical missing value replacement
`NORMALIZATION_SHIFT`	Normalization shift value
`NORMALIZATION_SCALE`	Normalization scale value

Parent topic: Model Detail Views

4.9.26 Model Detail Views for Exponential Smoothing

Model detail views specific to Exponential Smoothing (ESM) include information about the model output, global information about the model, and views that support time series regression.

These are the available model views for ESM:

Model Details	Description
`DM$VG`model_name	Global Name-Value Pairs
`DM$VP`model_name	Exponential Smoothing Forecast
`DM$VS`model_name	Computed Settings
`DM$VW`model_name	Model Build Alerts
`DM$VR`model_name	Time Series Regression Build
`DM$VT`model_name	Time Series Regression Score

Exponential Smoothing Forecast view (DM$VPmodel_name) displays the outcome of an ESM model. The output contains a set of records, ordered by partition and CASE_ID, that include the columns given in the Exponential Smoothing Model Output table. CASE_ID identifies the value's position in the time series. The user-specified CASE_ID can be a type that represents a numerical or datetime value. For each unique value of PARTITION, a distinct exponential smoothing model is built. The VALUE column for each PARTITION represents the observed or accumulated value of the target at that point in the sequence. The PREDICTION column is the forecast one step ahead at that point in the sequence. Backcasts are predictions that fall inside the range of the input data. The sequence also includes a user-specified number of values beyond the range of the input data. The VALUE column is NULL for any sequence value outside the range of input, and PREDICTION column is the model forecast for that sequence value. Lower and upper boundaries of the forecasts are denoted by the LOWER and UPPER columns. For backcasts, LOWER and UPPER are NULL. The bounds are based on a confidence interval that the user sets for the prediction.

Table 4-83 Exponential Smoothing Forecast View

Name	Description
`PARTITION`	Partition name in a partitioned model
`CASE_ID`	Sequence identifier (datetime or number type)
`VALUE`	Observed or accumulated value
`PREDICTION`	Backcast or Forecast value
`UPPER`	Upper bound of the forecast
`LOWER`	Lower bound of the forecast

Global Name-Value Pairs view (DM$VGmodel_name) includes the model's global information as well as the estimated smoothing constants, estimated initial state, and global diagnostic measures.

Depending on the type of model, the global diagnostics include some or all of the following for Exponential Smoothing.

Table 4-84 Global Name-Value Pairs View for ESM

Name	Description
`–2 LOG-LIKELIHOOD`	Negative log-likelihood of model
`ALPHA`	Smoothing constant
`AIC`	Akaike information criterion
`AICC`	Corrected Akaike information criterion
`AMSE`	Average mean square error over user-specified time window
`BETA`	Trend smoothing constant
`BIC`	Bayesian information criterion
`GAMMA`	Seasonal smoothing constant
`INITIAL LEVEL`	Model estimate of value one time interval prior to start of observed series
`INITIAL SEASON i`	Model estimate of seasonal effect for season i one time interval prior to start of observed series
`INITIAL TREND`	Model estimate of trend one time interval prior to start of observed series
`MAE`	Model mean absolute error
`MSE`	Model mean square error
`PHI`	Damping parameter
`STD`	Model standard error
`SIGMA`	Model standard deviation of residuals

Time series regression expands the features that can be included in a time series model and, possibly, increases forecast accuracy. Backcasts and forecasts of time series correlated to the "target" series of interest are included in the build and score views. The build and score views can be fed into a regression technique like Generalized Linear Model.

The Time Series Regression Build view (DM$VRmodel_name) depicts the schema for the build view. Each predictor series will have its own column. There can be a maximum of 20 predictor series in the build and score views. The names of the columns are obtained from the EXSM_SERIES_LIST setting.

Table 4-85 Time Series Regression Build View

Name	Description
`PARTITION`	Partition name in a partitioned model
`CASE_ID`	Sequence identifier (datetime or number type)
`target series name`	Observed or accumulated value of target series
`DM$target series`	Backcasted value of target series
`DM$predictor series column name`	Backcasted value of predictor series column. A maximum of 20 predictor series columns can be used.

The Time Series Regression Score view (DM$VTmodel_name) shows the schema for the score view. The schema is the same as in the build view, but the values in the target series name column are NULL because the future has not yet been observed.

Table 4-86 Time Series Regression Score View

Name	Description
`PARTITION`	Partition name in a partitioned model
`CASE_ID`	Sequence identifier (datetime or number type)
`target series name`	`NULL`s, because the future values of the target series have not been observed
`DM$target series`	Forecasted value of target series
`DM$predictor series column name`	Forecasted value of predictor series column name. A maximum of 20 predictor series columns can be used.

Related Topics

Parent topic: Model Detail Views

4.9.27 Model Detail Views for Text Features

The model details view for text features is DM$VXmodel_name.

The text feature view DM$VXmodel_name describes the extracted text features if there are text attributes present. The view has the following schema:

Name                                Type
 --------------            ---------------------
 PARTITION_NAME                     VARCHAR2(128)
 COLUMN_NAME                        VARCHAR2(128)
 TOKEN                              VARCHAR2(4000)
 DOCUMENT_FREQUENCY                 NUMBER

Table 4-87 Text Feature View for Extracted Text Features

Column Name	Description
`PARTITION_NAME`	A partition in a partitioned model to retrieve details
`COLUMN_NAME`	Name of the identifier column
`TOKEN`	Text token which is usually a word or stemmed word
`DOCUMENT_FREQUENCY`	A measure of token frequency in the entire training set

Parent topic: Model Detail Views

4.9.28 Model Detail Views for ONNX Models

You can view the details of an embedding model using the model detail views. The names of the views begin with DM$V.

This section lists the model detail views for embedding models.

DM$VJ Model Detail View
The DM$VJ<model-name> returns a single row containing a JSON object in one column that contains user-specified metadata of the model.
DM$VM Model Detail View
The DM$VM<model-name> view reports information extracted from the metadata of the imported ONNX model and its input or output tensors.
DM$VP Model Detail View
The DM$VP<model-name> view displays information extracted from parsing the JSON metadata. The view presents the JSON metadata of the model, including both explicitly declared properties and system-assigned default values for undeclared ones.

Parent topic: Model Detail Views

4.9.28.1 DM$VJ Model Detail View

The DM$VJ<model-name> returns a single row containing a JSON object in one column that contains user-specified metadata of the model.

The view has the following columns:

Name					   Null?    Type
------------------------ -------- ----------------------------
 METADATA					       CLOB

Column Name	Description
`METADATA`	It is a `CLOB` containing the user-specified metadata of the embedding model in JSON format.

The following table describes the output of the DM$VJ<modle_name> view of an embedding model.

Name	Value
`METADATA`	The JSON that was specified to the `IMPORT_ONNX_MODEL` call for importing the model.

The following example displays the output of an embedding model. The name of the model is doc_model:

select * from DM$VJdoc_model;

The output is as follows:


METADATA
--------------------------------------------------------------------------------
{"function":"embedding","embeddingOutput":"embedding","input":{"input":["DATA"]}}

Parent topic: Model Detail Views for ONNX Models

4.9.28.2 DM$VM Model Detail View

The DM$VM<model-name> view reports information extracted from the metadata of the imported ONNX model and its input or output tensors.

The view has the following columns:

Name                               Type
---------------------------------- ---------------------------
NAME                     	   VARCHAR2(4000)         
VALUE                              VARCHAR2(4000)

Table 4-88

Column Name	Description
`NAME`	The name of the metadata extracted from the ONNX model.
`VALUE`	Indicates a value for the metadata name

The following table describes the output of the DM$VM<model_name> view of an embedding model.

Name	Value
Producer Name	Name of the tools that generated the ONNX files
Graph Name	Name of the ONNX graph
Graph Description	Description given to the model
Version	Version of the model
Input	Describes the model input mapping
Output	Reports the vector information with dimension and value type

The following example displays the output of an embedding model. The name of the model is DOC_MODEL:

select * from DM$VMdoc_model;

The following is the output:


NAME                                     VALUE
---------------------------------------- ----------------------------------------
Producer Name                            onnx.compose.merge_models
Graph Name                               g_8_main_graph_main_graph
Graph Description                        Graph combining g_8_main_graph and main_
                                         graph
                                         g_8_main_graph
 
 
 
                                         main_graph
 
 
Version                                  1
Input[0]                                 input:string[1]
Output[0]                                embedding:float32[?,384]
 
6 rows selected.

Related Topics

https://github.com/onnx/onnx/blob/main/docs/IR.md

Parent topic: Model Detail Views for ONNX Models

4.9.28.3 DM$VP Model Detail View

The DM$VP<model-name> view displays information extracted from parsing the JSON metadata. The view presents the JSON metadata of the model, including both explicitly declared properties and system-assigned default values for undeclared ones.

The reported properties are specific to the machine learning model and match the mandatory and optional fields of the JSON metadata.

The view has the following columns:

Name                               Type
---------------------------------- ---------------------------
NAME                     	   VARCHAR2(4000)         
VALUE                              VARCHAR2(4000)

Column Name	Description
`NAME`	Displays the JSON parameters
`VALUE`	Indicates the value corresponding to the JSON parameter name value pair

Note that this information is already available in the ALL_MINING_MODEL_ATTRIBUTES view. The following example displays all the columns available to you in the DM$VPdoc_model view of an embedding model. In this example, doc_model is the name of the model.

select * from DM$VPdoc_model;

The output is as follows:


NAME                                     VALUE
---------------------------------------- ----------------------------------------
batching                                 False
embeddingOutput                          embedding

Parent topic: Model Detail Views for ONNX Models