Dimensionality reduction using genetic algorithm for improving accuracy in medical diagnosis

Abstract

The technological growth generates the massive data in all the fields. Classifying these high-dimensional data is a challenging task among the researchers. The high-dimensionality is reduced by a technique is known as attribute reduction or feature selection. This paper proposes a genetic algorithm (GA)-based features selection to improve the accuracy of medical data classification. The main purpose of the proposed method is to select the significant feature subset which gives the higher classification accuracy with the different classifiers. The proposed genetic algorithm-based feature selection removes the irrelevant features and selects the relevant features from original dataset in order to improve the performance of the classifiers in terms of time to build the model, reduced dimension and increased accuracy. The proposed method is implemented using MATLAB and tested using the medical dataset with various classifiers namely Näve Bayes, J48, and k-NN and it is evident that the proposed method outperforms other methods compared.