Beginners

For beginners, the best place to start is to use the SubmitX Visual Model Designer which comes with the “Autopilot” option. Beginners can upload a raw file, create new features,issue a learn request and instantly verify the predicted result.

Download the SubmitX Visual Model Designer from here, follow the installation instruction and instantly deploy your first machine-learning project.

Experts

Experts can extend the capability of the SubmitX ML Library by creating and plugging new algorithms, sub-classing existing algorithms, creating and extending feature engineering nodes.

public class JantricANNLearningEnv extends JantricLearningEnvBase
{
...
@Override
public HashMap<String, Double> learnFromData(ICJDataFrame dataFrame) throws JantricAttributeException 

{   
ICJMatrix _dataFrame = dataFrame.getDummyDataFrameMat(this.getDisEncod());			
ICJHiddenLayerManager m_hdnLyrMgr = (ICJHiddenLayerManager) ( _actionNeuron).getGraphMemoryAccess(ICJHiddenLayerManager.NODE_TYPE);
			m_hdnLyrMgr.setDataSize(_dataFrame.getNumRows());
	m_hdnLyrMgr.setNumberOfFeatures(_numberOfFeatures);m_hdnLyrMgr.setActiv("SINUSOID");
ICJMatrix _outpt = m_hdnLyrMgr.getLearnedANNOutput(false);	        				ICJMatrix _confusionMat = _testRegresand.getConfusionMatrix(_outpt);
double recall = _confusionMat.CJMatToCJBigMat().getAvgRecall().doubleValue();
double precision = _confusionMat.CJMatToCJBigMat().getAvgPrecision().doubleValue();
double f1Score = (2*precision*recall)/(recall+precision);