Using Weka s feature selection algorithm X, the features on current trunk, each colum is a feature vector on all instances, and each row is a part of the instance Y, the label of instances, in single column form: 1 2 3 4 5 ... a.E = w