I'm trying to build a classifier on JAVA using Weka API version 3.7. Firstly, i load the dataset with this portion of code:
InputStream fichier=new FileInputStream("E:\Participation\train_data.libsvm");
SVMLightLoader bdd= new SVMLightLoader();
bdd.setSource(fichier);
Instances dataset= new Instances(bdd.getDataSet());
System.out.println(dataset.toSummaryString());
the dataset is loaded successfuly. However, to apply the classifier (e.g: Naive Bayes) i need the class attribute which must be a Nominal type. So, how i can implement this filter on my dataset .
Thank You!
