Some Studies and Inferences on Pima Indian Diabetes Data

Abstract

Diabetes is a growing health condition and its screening is critical for early diagnosis. Because it is difficult to categorise individuals based on their stage of diabetes, type and treatments, it is an issue that needs a lot of attention. This project presents a study using Factor Analysis based on factor reduction that will help for diabetes diagnosis (type 2). The Pima dataset was used for the study. The discovery of the most influential factors was aided by factor reduction. The exploratory factor analysis (EFA) was used. After the analysis, the most influential factor combinations were determined to be pregnancies and age, BMI and skin thickness, Insulin and glucose. Three factor combinations were obtained. Pregnancy age factor, Body mass factor and Blood Glucose factor. Factor analysis worked well with the dataset and the most influential features that can be used for predicting diabetes.