NEIWDA with Hybrid Ensemble Model for Diagnosis of Diabetes Type-II

Abstract

Currently the accessibility of medical data increases gradually with the purpose is able to aid as valuable information for diagnosis by applying data mining apparatus. Diabetes is fundamentally the mainly general frequent disease in each and every one populace with age corporations. Many of the citizens belong to the Type-2 Diabetes Mellitus (T2DM) with the threat factors of T2DM are also higher between them.  The recent work machine learning based ensemble model is introduced for T2DM prediction. On the other hand, it doesn’t choose optimal features to increase the prediction results. To manage this issue, the proposed system designed Natural Exponential Inertia Weight Based Dragonfly Algorithm (NEIWDA) with hybrid machine learning based ensemble model for increasing the T2DM accuracy. Initially, Pima indians diabetes data is in use as an input. The optimal features are selected by Natural Exponential Inertia Weight based Dragonfly Algorithm (NEIWDA). Then the classification is performed using ensemble method.  In the ensemble approach, K-Nearest Neighbour (K-NN), Adaptive Network-based Fuzzy Inference System (ANFIS), Artificial Neural Networks (ANN), and Support Vector Machine (SVM) classifiers are utilized. In Ensemble method the results of individual classifiers such as K-NN, ANN and SVM are fused together. Thus results increase the performance and the chances of reducing misclassification, this gives an improved performance to the overall classification performance. Ensemble technique uses a majority voting and provides us the get into results. The outcome described that the designed method gives improved performance matched with the previous methods in terms of accuracy, precision, recall as well as f-measure.