An efficient metrics based self‐adaptive design model by multiobjective gray wolf optimization with extreme learning machine for autonomic computing system application

Abstract

AbstractAutonomic computing refers to the capability of the software system to monitor environmental changes by itself. The primary objective of the proposed approach is to identify the quality of software components concerning complexity metrics. In this article, extreme learning machine (ELM) supported multiobjective gray wolf optimization (MOGWO) is proposed to predict the quality products. The proposed methodology considers four component‐based system metrics such as failure response time, throughput rate, interface surface consistency (ISC) and bounded interface complexity metrics (BICM) to resolve the complexity issues. The ELM is utilized in this article to conduct a prediction in the complexity metrics on the context‐awareness self‐adaptive autonomous system application. MOGWO is then utilized to optimize the cost metric. The evaluation results are compared with the existing approaches such as ELM, ANN, and ANN‐GWO. From the simulations, the overall accuracy rate of the prediction model is obtained to be 95%.