Prediction of Diabetes using Adaptive Neuro Fuzzy Inference System (ANFIS)

Abstract

In this work a systematic architecture and procedure of Adaptive Neuro Fuzzy Inference System (ANFIS) is presented for data mining classification model to predict the Pima diabetes dataset and compare the model errors with Artificial Neural Network (ANN) model. In the Proposed ANFIS model, an input-output data has been mapped on both fuzzy if-then rules and stipulated input-output data pairs. Then the architecture of ANFIS model is employed and the results are compared. It was found that the proposed ANFIS model outperforms ANN model in terms of prediction accuracy.