PREDICTION ACCURACY COMPARISON OF PREDICTIVE MODELS USING MACHINE LEARNING FOR DIABETES DATA SET

Abstract

Diagnosis of Diabetes disease at beginning stage is important for healthier treatment. In today’s scenario equipments like sensors are used for discovery of infections. Accurate classification techniques are necessary for automatic detection of disease samples. this study utilizes data mining techniques for classification of Diabetes patients. Five algorithms (Logistic Regression and Artificial Neural Network, SVM, Random forest) were implemented for classification using R platform. Classification and prediction of medical datasets poses real challenges in Data Mining. To deal with these challenges Logistic Regression (LR) and Artificial Neural Network (ANN) SVM , Random Forest are commonly used. LR enables us to examine the relationship between a categorical outcome and a set of descriptive variables. LR explains that there can be one or more self-governing variables that can establish the problem outcome. ANN resembles the human brain and here the information is processed by simple elements called neurons and signals are transmitted between the neuron From the experimental results it is identified that for Diabetes dataset NN with 10 fold using percentage split prediction correctness of 84.52% is achieved.