The Impact of Predictor Variables for Detection of Diabetes Mellitus Type-2 for Pima Indians

Abstract

Diabetes Mellitus type-2 is one of the diseases of a modern age treated as a serious illness due to its symptoms in later stages, consequences if left untreated and its complexity in terms of detection, diagnosis, and prognosis widely spread among the Pima Indian population. The process of detecting the diabetes will require analysis of the data, processing, extracting portions of data into a set for training, testing and validation sets. Then applying several different machine learning algorithms, train a model, check the performance of the trained model and iterate with other algorithms until we find the most performant for our type of data. The goal of this research is to investigate which algorithm gives best results in terms of detecting the existing disease as well as predicting the possibility of getting one in the future, based on the diagnostic measurements of the patient. For this matter, MATLAB software will be used.