9.5 Automatic Data Preparation
Oracle Machine Learning for Python supports Automatic Data Preparation (ADP) and user-directed general data preparation.
The PREP_*
settings enable you to request fully
automated (ADP) or manual data preparation. By default, ADP is enabled
(PREP_AUTO_ON
) . When performed manually, data preparation
requirements of each algorithm must be addressed
When you enable ADP, the model uses heuristics to transform the build
data according to the requirements of the algorithm. Instead of ADP, you can request
that the data be shifted and/or scaled with the PREP_SCALE_*
and
PREP_SHIFT_*
settings. The transformation instructions are
stored with the model and reused whenever the model is applied. The model settings
can be viewed in USER_MINING_MODEL_SETTINGS
.
PREP_* Settings
The values for the PREP_*
settings are described in the
following table.
Table 9-2 title
Setting Name | Setting Value | Description |
---|---|---|
PREP_AUTO |
|
This setting enables fully automated data preparation. The default is |
PREP_SCALE_2DNUM |
|
This setting enables scaling data preparation for
two-dimensional numeric columns.
|
PREP_SCALE_NNUM |
|
This setting enables scaling data preparation for
nested numeric columns. |
PREP_SHIFT_2DNUM |
|
This setting enables centering data preparation for
two-dimensional numeric columns.
|
See Also: