Predictive Design to Analyze Diabetes using Machine Learning Classifier

Authors

  • Merlin Livington, L.Sujihelen, C.Senthilsingh

Abstract

Today’s advent in the medical industry have given numerous chances to improve the quality of detection and reporting the diseases at the early stages for a better diagnosis. Diabetes mellitus is one such disease which is predominant among a global population which ultimately leads to blindness and death in some cases. The model proposed in this system attempts to design and deliver an intelligent solution for predicting diabetes in the early stages and address the problem of late detection and diagnosis. Intensive research is carried out in many tropical countries for automating this process through a machine learning model. The accuracy of machine learning algorithms is more than satisfactory in detection of Type 2 diabetes from the dataset of PIMA Indians Diabetes Dataset. An additional feature of hereditary factor is implemented to the existing multiple objective fuzzy classifiers. The proposed model has improved the accuracy to 83% in the training and tested datasets when compared to existing method.                                                                                                                              

Keywords-Fuzzy classifiers, Diabetes Dataset, machine learning, Gestational diabetes

Published

2020-12-11

Issue

Section

Articles