Implementation of Novel Algorithm in the Rule Based Expert System for Classification Accuracy

Abstract

This research paper objective is to provide analysis of classifying the data set using novel naïve classification algorithm. Here this algorithm applied on two dataset which are the first one is training data set and other one is test data set, and this experiment attempts to compare the classification and accuracy of the proposed algorithm with the two dataset . Several constraints used for analytical purpose which are classification accuracy, True Positive Rate, False positive Rate sensitivity, Precision, Recall and f-measure using confusion matrix. Finally results are given in the tabular form to facilitate comparative analysis. From this study it is found that this algorithm given good result for both train as well as test data set but more accurate for train set. Section I discussed the general approach of classification, section II discussed Literature review, section III discussed the proposed work about the performance of train and test dataset, result and discussion briefed in the section IV and conclusion in section V