Genetic algorithm based feature selection and MOE Fuzzy classification algorithm on Pima Indians Diabetes dataset

Diabetes Mellitus is a dreadful disease characterized by increased levels of glucose in the blood, termed as the condition of hyperglycemia. As this disease is prominent among the tropical countries like India, an intense research is being carried out to deliver a machine learning model that could learn from previous patient records in order to deliver smart diagnosis. This research work aims to improve the accuracy of existing diagnostic methods for the prediction of Type 2 Diabetes with machine learning algorithms. The proposed algorithm selects the essential features from the Pima Indians Diabetes Dataset with Goldberg's Genetic algorithm in the pre-processing stage and a Multi Objective Evolutionary Fuzzy Classifier is applied on the dataset. This algorithm works on the principle of maximum classifier rate and minimum rules. As a result of feature selection with GA the number of features is reduced to 4 from 8 and the classifier rate is improved to 83.0435 % with NSGA II in training rate of 70% and 30% testing.