trainf_ocr_class_mlpT_trainf_ocr_class_mlpTrainfOcrClassMlpTrainfOcrClassMlptrainf_ocr_class_mlp (Operator)

Name

trainf_ocr_class_mlpT_trainf_ocr_class_mlpTrainfOcrClassMlpTrainfOcrClassMlptrainf_ocr_class_mlp — Train an OCR classifier.

Signature

trainf_ocr_class_mlp( : : OCRHandle, TrainingFile, MaxIterations, WeightTolerance, ErrorTolerance : Error, ErrorLog)

Description

trainf_ocr_class_mlptrainf_ocr_class_mlpTrainfOcrClassMlpTrainfOcrClassMlpTrainfOcrClassMlptrainf_ocr_class_mlp trains the OCR classifier OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle with the training characters stored in the OCR training files given by TrainingFileTrainingFileTrainingFileTrainingFiletrainingFiletraining_file. The training files must have been created, e.g., using write_ocr_trainfwrite_ocr_trainfWriteOcrTrainfWriteOcrTrainfWriteOcrTrainfwrite_ocr_trainf, before calling trainf_ocr_class_mlptrainf_ocr_class_mlpTrainfOcrClassMlpTrainfOcrClassMlpTrainfOcrClassMlptrainf_ocr_class_mlp.

The remaining parameters have the same meaning as in train_class_mlptrain_class_mlpTrainClassMlpTrainClassMlpTrainClassMlptrain_class_mlp and are described in detail with train_class_mlptrain_class_mlpTrainClassMlpTrainClassMlpTrainClassMlptrain_class_mlp. A regularization of the OCR classifier and an automatic determination of the regularization parameters (see set_regularization_params_ocr_class_mlpset_regularization_params_ocr_class_mlpSetRegularizationParamsOcrClassMlpSetRegularizationParamsOcrClassMlpSetRegularizationParamsOcrClassMlpset_regularization_params_ocr_class_mlp) is taken into account during the training. Furthermore, if a rejection class has been specified using set_rejection_params_ocr_class_mlpset_rejection_params_ocr_class_mlpSetRejectionParamsOcrClassMlpSetRejectionParamsOcrClassMlpSetRejectionParamsOcrClassMlpset_rejection_params_ocr_class_mlp, before the actual training the samples for the rejection class are generated.

Please note that training characters that have no corresponding class in the classifier OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle are discarded.

Execution Information

Multithreading type: reentrant (runs in parallel with non-exclusive operators).
Multithreading scope: global (may be called from any thread).
Automatically parallelized on internal data level.

This operator modifies the state of the following input parameter:

OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle

During execution of this operator, access to the value of this parameter must be synchronized if it is used across multiple threads.

Parameters

OCRHandleOCRHandleOCRHandleOCRHandleOCRHandleocrhandle (input_control, state is modified) ocr_mlp → (handle)

Handle of the OCR classifier.

TrainingFileTrainingFileTrainingFileTrainingFiletrainingFiletraining_file (input_control) filename.read(-array) → (string)

Names of the training files.

Default value: 'ocr.trf' "ocr.trf" "ocr.trf" "ocr.trf" "ocr.trf" "ocr.trf"

File extension: .trf, .otr

MaxIterationsMaxIterationsMaxIterationsMaxIterationsmaxIterationsmax_iterations (input_control) integer → (integer)

Maximum number of iterations of the optimization algorithm.

Default value: 200

Suggested values: 20, 40, 60, 80, 100, 120, 140, 160, 180, 200, 220, 240, 260, 280, 300

WeightToleranceWeightToleranceWeightToleranceWeightToleranceweightToleranceweight_tolerance (input_control) real → (real)

Threshold for the difference of the weights of the MLP between two iterations of the optimization algorithm.

Default value: 1.0

Suggested values: 1.0, 0.1, 0.01, 0.001, 0.0001, 0.00001

Restriction: WeightTolerance >= 1.0e-8

ErrorToleranceErrorToleranceErrorToleranceErrorToleranceerrorToleranceerror_tolerance (input_control) real → (real)

Threshold for the difference of the mean error of the MLP on the training data between two iterations of the optimization algorithm.

Default value: 0.01

Suggested values: 1.0, 0.1, 0.01, 0.001, 0.0001, 0.00001

Restriction: ErrorTolerance >= 1.0e-8

ErrorErrorErrorErrorerrorerror (output_control) real → (real)

Mean error of the MLP on the training data.

ErrorLogErrorLogErrorLogErrorLogerrorLogerror_log (output_control) real-array → (real)

Mean error of the MLP on the training data as a function of the number of iterations of the optimization algorithm.

Example (HDevelop)

* Train an OCR classifier
read_ocr_trainf_names ('ocr.trf', CharacterNames, CharacterCount)
create_ocr_class_mlp (8, 10, 'constant', 'default', CharacterNames, 80, \
                      'none', 81, 42, OCRHandle)
trainf_ocr_class_mlp (OCRHandle, 'ocr.trf', 100, 1, 0.01, Error, ErrorLog)
write_ocr_class_mlp (OCRHandle, 'ocr.omc')

Result

If the parameters are valid, the operator trainf_ocr_class_mlptrainf_ocr_class_mlpTrainfOcrClassMlpTrainfOcrClassMlpTrainfOcrClassMlptrainf_ocr_class_mlp returns the value TRUE. If necessary, an exception is raised.

trainf_ocr_class_mlptrainf_ocr_class_mlpTrainfOcrClassMlpTrainfOcrClassMlpTrainfOcrClassMlptrainf_ocr_class_mlp may return the error 9211 (Matrix is not positive definite) if Preprocessing = 'canonical_variates'"canonical_variates""canonical_variates""canonical_variates""canonical_variates""canonical_variates" is used. This typically indicates that not enough training samples have been stored for each class. In this case we recommend to change Preprocessing to 'normalization'"normalization""normalization""normalization""normalization""normalization". Another solution can be to add more training samples.