TrainModel Method of the Trainer Object

TrainModel method of the Trainer object in the ABBYY FineReader Engine API — trains the classifier on the supplied TrainingData using the configured TrainingParams and ValidationParams.

Trains the classifier on the input data set. The training and cross-validation parameters should be set up in the properties of the Trainer object.

Model training and classification will be performed in sequential mode in Linux and Windows, regardless of the IMultiProcessingParams::MultiProcessingMode value.

Syntax

C++

HRESULT TrainModel(
  ITrainingData*     Objects,
  ITrainingResults** Result
);

C#

ITrainingResults TrainModel( ITrainingData Objects );

Visual Basic .NET

Function TrainModel(Objects As ITrainingData) As ITrainingResults

Parameters

Objects

[in] The categorized data set (as a TrainingData object), on which the classification model should be trained. At least two categories must be represented in the data set.

Result

[out, retval] A pointer to the ITrainingResults* pointer variable that receives the interface pointer of the output TrainingResults object. The results of training include the classification model and the validation results. Currently you can only train a classifier of one type at a time, therefore the collection will contain one TrainingResult element.

Return values

This method has no specific return values. It returns the standard return values of ABBYY FineReader Engine functions.

See also

Trainer