Report

Declaration: TRndForest.Report [clsix: integer]: TRFReport;
The readonly array property Report returns various error and reliability measures of the model. The parameter clsix specifies the class index and may assume values between 1 and NClasses.

Each element of the array contains the following charateristic parameters for the corresponding class:

RelClsError Training Set - Relative classification error (percent of incorrectly classified cases)
AvgCE Training Set - Average cross-entropy (in bits per element)
RMSError Training Set - root mean square error when estimating posterior probabilities
AvgError Training Set - Average error when estimating posterior probabilities
AvgRelError Training Set - Average relative error when estimating posterior probability of belonging to the correct class
OOBRelClsError OOB Test - Relative classification error (percent of incorrectly classified cases)
OOBAvgCE OOB Test - Average cross-entropy (in bits per element)
OOBRMSError OOB Test - root mean square error when estimating posterior probabilities
OOBAvgError OOB Test - Average error when estimating posterior probabilities
OOBAvgRelError OOB Test - Average relative error when estimating posterior probability of belonging to the correct class