Adaptive Bayes Network Settings
Note: The Adaptive Bayes Network (ABN) algorithm is deprecated in Oracle Data Mining 11g. Oracle Data Miner does not support building ABN models; you can view ABN models created using the Oracle Data Mining 11g APIs.
The default settings are designed so that they should work well
for most cases.
First select the model type;
Model Type is one of the following:
- Single Feature, the default
- Multi Feature
- Naive Bayes
Note: Single Feature is the only model type that provides rules.
See Adaptive Bayes Network Algorithm for more information about the model types.
The values that you can change and their defaults depend on the Model Type that you select.
You can change the following for Single Feature:
- Maximum Predictors: A feature selection mechanism that can provide a substantial performance improvement. The actual number of predictors is the minimum of this value and the number of active predictors in the model. If the value is less than the number of active predictors in the model, the predictors are chosen in accordance with their rank. The default is 25.
- Run Time Limit: The number of minutes for the build operation; the default is to specify no time limit.
You can change the following for Multi Feature:
- Maximum Predictors: A feature selection mechanism that can provide a substantial performance improvement, especially in the instance of wide training tables. The actual number of predictors is the minimum of this value and the number of active predictors in the model. If the value is less than the number of active predictors in the model, the predictors are chosen in accordance with their rank. The default is 25.
- Run Time Limit: The number of minutes for the build operation; the default is to specify no time limit.
- Naive Bayes Predictors: The maximum number of attributes to consider during the build process. The default is 10. The larger the number, the slower the build. This number must be less than the number of features.
You can change the following for Naive Bayes:
- Naive Bayes Predictors: The maximum number of attributes to consider during the build process (in rank order MDL correlation to target). The default is 10. The larger the number, the slower the build. This number must be less than the number of features.
Click OK to continue.
Copyright © 2006, 2008, Oracle. All rights
reserved.