Classifier allows the user to classify between different classes based on the features acquired. The goals and applications of different classifiers are different. As the feature selection is one of the important criteria. In this paper we introduce a method of ranking the features of one class with respect to another and it tells the user that in the training set which feature has higher ranking among the other. So this method tells which feature is insignificant in certain classes and it can be ruled out. The classification can be made so easily as for some cases, certain features creates confusion in the classifier and wrong interpretations are also occurs. In the training set, if a new data is given as input and this method able to tell the user that the features has a variation with respect to training data set and the feature ranking is calculated. This method automatically ranks the feature and feature selection can be made easier. So we can able to interpret from the significant and insignificant features.
S. Padmavathi and Krishnan, S., “Feature Ranking Procedure for Automatic Feature Extraction”, in International conference on Signal Processing, Communication, Power and Embedded System (SCOPES)-2016, 2016.