Predicting Cardiovascular Disease In Type 2 Diabetic Patients By Feature Selection Using Particle Swarm Optimization

Type 2 Diabetes is serious global health problem for most countries and it is the most common disease nowadays in all populations and in all age groups. An efficient predictive modeling is required for medical researchers and practitioners to improve the prediction accuracy of the classification methods. The aim of this research was to identify significant factors influencing type 2 diabetes control with CVD risk factors, by applying particle swarm optimization feature selection system to improve prediction accuracy and knowledge discovery. Proposed system consists of four major steps such as preprocessing and dimensionality reduction of type 2 diabetes with CVD factors, Attribute Value Measurement, Feature Selection, and Hybrid Prediction Model. In proposed methods the preprocessing and dimensionality reduction of the patients records is performed by using Kullback Leiber Divergence(KLD)-Principal component analysis (PCA) ,then attribute values measurement is performed using Fast Correlation-Based Filter Solution (FCBFS), feature selection is performed by using Particle Swarm Optimization (PSO)