Hybrid Adaptive Moth-Flame Optimization and Opposition-Based Learning with its Applications

ABSTRACT

This paper is devoted to improving the optimization capacities of moth-flame optimizer (MFO), which works based on a spiral operation mimicking the moths' navigation behavior and dynamic coefficients. However, such a basic version can be easily trapped in the local optima and is associated with unstable balance states between exploratory and exploitative cores. To mitigate the shortcoming of slow convergence and local stagnation, we have proposed AMFOOBL employing a diversification core: opposition-based learning (OBL) strategy and a new adaptive structure combined. This original adaptive mechanism is designed to reduce the number of flames around which agents update their positions for balancing the exploration and exploitation stages more effectively. The proposed algorithm's performance is evaluated on three experiments: Firstly, the quantitative results of 23 benchmark function tests show that AMFOOBL outperforms AMFO, followed by MFO, demonstrating the effectiveness of our new adaptive approach in terms of accuracy and convergence rate. Secondly, AMFOOBL is demonstrated on Multilayer Perceptron's structural realization and training compared with nine state-of-the-art algorithms. The simulation on eight datasets, including pattern classification and function approximation, reveals outstanding performance in the AMFOOBL-based trainer concerning classification accuracy and test error. Moreover, AMFOOBL is verified on multilevel thresholding segmentation of color images, supporting its excellent robustness in both minimization and maximization problems. This paper's finding suggests that AMFOOBL is a superior algorithm, and the developed evolutionary-enhanced MLP can be considered a useful tool.