A Grey Prediction Model for Prognosticating Diabetes Mellitus using Hybrid Artificial Bee Colony Optimization

Authors

  • M.Durgadevi, T.Veeramakali, K.Abinaya, S.Kiruthiga

Abstract

Diabetes mellitus is considered as the collection of metabolic disorders in humans that leads to major complications when they are not identified and treated. The process of prognosticating the evidence of diabetes in humans leads to the emergence of a significant amount of clinical and genetic data collected from the individuals. The effective and efficient process of detecting diabetes in humans must incorporate better feature selection processes and a potential optimization scheme for classification and prediction model. In this paper, A Grey Prediction Model enforcing Relief Oriented Feature selection and Hybrid Artificial Bee Colony with Monarchy Butterfly Optimization-based Classification Scheme (GPMROFS-HABCMBO-CS) is contributed for predominant prognostication of Diabetes Mellitus. In this proposed GPMROFS-HABCMBO-CS, initially the merits of Relief-based Feature set selection process are used for significant feature selection. Then, the process of feature set reduction and optimizations is facilitated using meta-heuristic Hybrid Artificial Bee Colony with Monarchy Butterfly Optimization process, such that Grey Model-based Prediction can be ensured for remarkable detection of diabetes mellitus from the collected datasets used for investigation. The experimental investigation of the proposed GPMROFS-HABCMBO-CS was conducted using Pima Indians Diabetes Database. The performance was evaluated in terms of Accuracy, Precision, F-Measure and Recall which confirmed a considerable enhancement on par with the benchmarked Diabetes Mellitus diagnosing prediction models in the literature.

Published

2020-11-01

Issue

Section

Articles