ENH: implement TrainRegression for ANN

Modules/Learning/Supervised/include/otbNeuralNetworkMachineLearningModel.h

@@ -178,11 +178,14 @@ protected:
   /** PrintSelf method */
   void PrintSelf(std::ostream& os, itk::Indent indent) const;
 
-  /** Train the machine learning model */
+  /** Train the machine learning model for classidication*/
   virtual void TrainClassification();
-  /** Predict values using the model */
+  /** Predict values using the model in classification mode*/
   virtual TargetSampleType PredictClassification(const InputSampleType& input, ConfidenceValueType *quality=NULL) const;
 
+  /** Train the machine learning model for regression*/
+  virtual void TrainRegression();
+
 private:
   NeuralNetworkMachineLearningModel(const Self &); //purposely not implemented
   void operator =(const Self&); //purposely not implemented

Modules/Learning/Supervised/include/otbNeuralNetworkMachineLearningModel.txx

@@ -130,7 +130,7 @@ void NeuralNetworkMachineLearningModel<TInputValue, TOutputValue>::LabelsToMat(c
     }
 }
 
-/** Train the machine learning model */
+/** Train the machine learning model for classification*/
 template<class TInputValue, class TOutputValue>
 void NeuralNetworkMachineLearningModel<TInputValue, TOutputValue>::TrainClassification()
 {
@@ -213,6 +213,44 @@ typename NeuralNetworkMachineLearningModel<TInputValue, TOutputValue>::TargetSam
   return target;
 }
 
+/** Train the machine learning model for regression*/
+template<class TInputValue, class TOutputValue>
+void NeuralNetworkMachineLearningModel<TInputValue, TOutputValue>::TrainRegression()
+{
+  //Create the neural network
+  const unsigned int nbLayers = m_LayerSizes.size();
+
+  if ( nbLayers == 0 )
+    itkExceptionMacro(<< "Number of layers in the Neural Network must be >= 3")
+
+  cv::Mat layers = cv::Mat(nbLayers, 1, CV_32SC1);
+  for (unsigned int i = 0; i < nbLayers; i++)
+    {
+    layers.row(i) = m_LayerSizes[i];
+    }
+
+  m_ANNModel->create(layers, m_ActivateFunction, m_Alpha, m_Beta);
+
+  //convert listsample to opencv matrix
+  cv::Mat samples;
+  otb::ListSampleToMat<InputListSampleType>(this->GetInputListSample(), samples);
+
+  cv::Mat matOutputANN;
+  otb::ListSampleToMat<TargetListSampleType>(this->GetTargetListSample(), matOutputANN);
+
+  CvANN_MLP_TrainParams params;
+  params.train_method = m_TrainMethod;
+  params.bp_dw_scale = m_BackPropDWScale;
+  params.bp_moment_scale = m_BackPropMomentScale;
+  params.rp_dw0 = m_RegPropDW0;
+  params.rp_dw_min = m_RegPropDWMin;
+  CvTermCriteria term_crit = cvTermCriteria(m_TermCriteriaType, m_MaxIter, m_Epsilon);
+  params.term_crit = term_crit;
+
+  //train the Neural network model
+  m_ANNModel->train(samples, matOutputANN, cv::Mat(), cv::Mat(), params);
+}
+
 template<class TInputValue, class TOutputValue>
 void NeuralNetworkMachineLearningModel<TInputValue, TOutputValue>::Save(const std::string & filename,
                                                                         const std::string & name)