Java: Data Science Made Easy

Evaluating the model

The next step is to evaluate the model. The Evaluation class is used for this purpose. Its constructor takes the training set as input and the evaluateModel method performs the actual evaluation. The following code illustrates this using the testing dataset:

Evaluation evaluation = new Evaluation(trainingInstances); evaluation.evaluateModel(mlp, testingInstances); 

One simple way of displaying the results of the evaluation is using the toSummaryString method:

System.out.println(evaluation.toSummaryString()); 

This will display the following output:

Correctly Classified Instances 73 98.6486 %Incorrectly Classified Instances 1 1.3514 %Kappa statistic 0.9824Mean absolute error 0.0177Root mean squared error 0.076 Relative ...