An ensemble approach to meta-heuristic algorithms: Comparative analysis and its applications

Abstract

In this paper, we intend to propose an ensemble optimization algorithm based on Follow The Leader (FTL), Multi-verse Optimizer (MVO), and Salp Swarm Algorithm (SSA) to solve constrained optimization problems. The FTL, MVO, and SSA are swarm-based algorithms that update their particle position using a selection approach. Less number of control parameters and a common selection approach make these algorithms suitable for hybridization. In this work, combinations of FTL, MVO, and SSA algorithms such as FTL_MVO, FTL_SSA, MVO_SSA, and FTL_MVO_SSA have been proposed to solve different optimization problems. The proposed ensemble optimization algorithms have been compared with base optimization algorithms on forty-eight unimodal and multimodal benchmark functions. The ensemble model has achieved significant performance improvement over base FTL, MVO, and SSA. Moreover, these algorithms have been tested on six well-known constrained optimization problems to benchmark their performance over real-world applications. Finally, the comparison with classical optimization algorithms reveals the efficacy of the proposed models.