A visualization tool for observing metaheuristic algorithm: Epanel
Yao–Hsin Chou, Shu–Yu Kuo, Wei-Jie Lai, Wang-Bin Lai · 2017
Metaheuristic algorithms are important methods to solve many complicated optimization problems. However, there are no significant studies that present an application to analyze metaheuristic algorithms or a platform to help beginners to learn about. For this reason, we design a useful tool with a graphical user interface, EPanel, to dynamically display the process of metaheuristic algorithms. EPanel displays the value of a function in one or two dimensions by using different colors. It helps users understand the distribution of the solution space, such as unimodal, multimodal, and narrow valley. EPanel then shows the position of every particle above the screen of a function in each iteration and displays the complete optimization process of the algorithm by animation. Implemented with JAVA, EPanel is able to be executed in different operating systems. EPanel helps researchers to analyze each step of the algorithm and to observe how the algorithm achieves exploratory and exploitation. By using EPanel, research can further develop or improve algorithms efficiently. Beginners can learn about metaheuristic algorithms through EPanel. They can see popular algorithms, such as Particle Swarm Optimization and Artificial Bee Colony, work in each movement. EPanel is a useful application and helps not only researchers but also beginners.