Oracle Machine Learning Product and Feature Matrix
Learn about supported algorithms and features for OML products.
Supervised Algorithms
The following table displays the supported supervised techniques and algorithms across Oracle Machine Learning components.
| Technique | Algorithm | OML4SQL | OML4R 2.0 | OML4R 1.5.1 ORE 1.5.1 | OML4Py 2.1 | OML4Py 2.0 | OML4Py 1.0 | Oracle Data Miner | AutoML UI |
|---|---|---|---|---|---|---|---|---|---|
| Classification | Decision Tree | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Explicit Semantic Analysis (ESA) | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Generalized Linear Model (GLM) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Ridge Regression | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Naive Bayes | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Logistic Regression | ✓ through GLM |
✓ through GLM |
✓ through GLM |
✓ | ✓ | ✓ through GLM |
✓ through GLM |
✓ | |
| Neural Network | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | - | ✓ | |
| Random Forest | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | - | ✓ | |
| Support Vector Machine (SVM) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Extreme Gradient Boosted Trees (XGBoost) | ✓ (21c) |
✓ (21c) |
- | ✓ | ✓ | - | - | - | |
| Regression | Generalized Linear Model (GLM) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Linear Regression Model | ✓ through GLM |
✓ through GLM |
✓ through GLM |
✓ | ✓ | ✓ through GLM |
✓ through GLM |
✓ | |
| Neural Network | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | - | ✓ | |
| Ridge Regression | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Support Vector Machine (SVM) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Extreme Gradient Boosted Trees (XGBoost) | ✓ (21c) |
✓ (21c) |
- | ✓ | ✓ | - | - | - | |
| Time Series | Simple exponential smoothing | ✓ | ✓ | - | ✓ | ✓ | - | - | - |
| Simple exponential smoothing with multiplicative error | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Holt linear exponential smoothing | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Exponential smoothing with multiplicative trend | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Exponential smoothing with multiplicative damped trend | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Exponential smoothing with additive seasonality, but no trend | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Exponential smoothing with multiplicative seasonality, but no trend | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Holt-Winters triple exponential smoothing, additive trend, multiplicative seasonality | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Holt-Winters multiplicative exponential smoothing with damped trend, additive trend, multiplicative seasonality | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Holt-Winters additive exponential smoothing, additive trend, additive seasonality | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Holt-Winters additive exponential smoothing with damped trend, additive trend, additive seasonality | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Holt-Winters multiplicative exponential smoothing with multiplicative trend, multiplicative trend, multiplicative seasonality | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
| Holt-Winters multiplicative exponential smoothing with damped multiplicative trend, multiplicative trend, multiplicative seasonality | ✓ | ✓ | - | ✓ | ✓ | - | - | - | |
|
Multiple Time Series (Supporting Time Series Regression) |
✓ (26ai) |
- | - | - | - | - | - | - | |
| Attribute Importance | Minimum Description Length | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | - |
| Random Forest | ✓ | ✓ | - | ✓ | ✓ | ✓ | - | - | |
| Survival Analysis | Extreme Gradient Boosted Trees (XGBoost) | ✓ (26ai) |
- | - | - | - | - | - | - |
Unsupervised Algorithms
The following table displays the supported unsupervised techniques and algorithms across Oracle Machine Learning components.
| Technique | Algorithm | OML4SQL | OML4R 2.0 | OML4R 1.5.1 ORE 1.5.1 | OML4Py 2.1 | OML4Py 2.0 | OML4Py 1.0 | Oracle Data Miner 19c |
|---|---|---|---|---|---|---|---|---|
| Clustering | Expectation Maximization (EM) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Gaussian Mixture Models | ✓ through EM |
✓ through EM |
✓ through EM |
✓ | ✓ | ✓ through EM |
✓ through EM |
|
| Hierarchical k-Means | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Orthogonal Partitioning (O-Cluster) | ✓ | ✓ | ✓ | ✓ | ✓ | - | ✓ | |
|
Feature Extraction |
Explicit Semantic Analysis (ESA) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Non-Negative Matrix Factorization (NMF) | ✓ | ✓ | ✓ | ✓ | ✓ | - | ✓ | |
| Principal Component Analysis (PCA) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Singular Value Decomposition (SVD) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Low Rank Matrix Factorization | ✓ through SVD |
- | - | - | - | - | - | |
|
Association Rules |
Apriori | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
|
Anomaly Detection |
1 Class Support Vector Machine | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| Multivariate State Estimation Technique (MSET-SPRT) | ✓ | - | - | - | - | - | - | |
| Expectation Maximization (EM) | ✓ (26ai) |
- | - | ✓ | - | - | - | |
| Attribute Importance | CUR Matrix Decomposition | ✓ | - | - | - | - | - | - |
Other Features
The following table displays other supported features across Oracle Machine Learning components.
| Area | Feature | OML4SQL | OML4R 2.0 | OML4Py 2.1 |
|---|---|---|---|---|
| Embedded Execution | API | NA |
Oracle AI Database: R, SQL, REST through ORDS Autonomous AI Database: R, SQL, REST |
Oracle AI Database: Python, SQL, REST through ORDS Autonomous AI Database: Python, SQL, REST |
| AutoML | API | - | - | ✓ |
| Pipeline | NA | - | ✓ | |
| Automatic Data Preparation | - | ✓ | ✓ | ✓ |
| Partitioned Models | - | ✓ | ✓ | ✓ |
| Nested Columns as predictors | - | ✓ | - | - |
| Import ONNX-format model from file | - | ✓ | - | ✓ |
| Import transformer from Hugging Face | - | - | - | ✓ |
|
Access 3rd-party packages |
API | - | ✓ | ✓ |
|
Vector data type support |
API | ✓ | - | ✓ |
Note:
Access 3rd-party packages:
-
Oracle AI Database: Users perform custom installation of supported R or Python packages.
-
Autonomous AI Database: Users install and manage packages using the Conda environment.
Vector data type support:
-
OML4SQL: See Vector Data Type Support.
-
OML4Py 2.1: See Manipulate database tables and views using familiar Python functions and syntax.
Supported OML4Py Classes and OML4R Functions for In-Database Algorithms
A list of supported OML4Py classes and OML4R functions is provided in the respective documentation. For more details, see:
-
For OML4Py 2.1: About Machine Learning Classes and Algorithms
-
For OML4R 2.0: Oracle Machine Learning for SQL Models Supported by Oracle Machine Learning for R
Parent topic: Machine Learning Overview