Improved Jaguar Algorithm with Tabu List to Solve Function Optimization Problem

• Main Authors: LAI, WANG-BIN, 賴王斌
• Other Authors: CHOU, YAO-HSIN
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/42062126405889230149

碩士 === 國立暨南國際大學 === 資訊工程學系 === 105 === Meta-heuristic is one of the popular research which is implemented to solve optimization problems in real life. In order to obtain the best solution in limited cost or time,traditional meta-Heuristics are devoted to balancing the capabilities of exploration and exploitation. For the sake of the purpose, traditional methods implement lots of weights (parameters) to balance two capabilities, and implement random variables or numerous population and generation to increase opportunities for finding a better solution. However, the searching mode of traditional methods might have some potential problems. Such as exploration and exploitation in traditional methods might restrict to each other. And implemented parameters shall be adjusted for different problems. Therefore, Jaguar Algorithm is designed in a new concept. We concentrate on exploitation before exploration. At first, proposed method tries its best to find the optimal solution in the arbitrary area. Then it focuses on jumping to better area based on the information of the history. Along the tendency of found areas to find out the place of the global optimum. The proposed method achieves strong capabilities of both exploitation and exploration with these features. Our idea comes from that the idea of territory is similar to Tabu Search Algorithm. The proposed method used Jaguar Algorithm territory’s information to prevent other jaguars or itself entering those searched areas again and reduce the evaluations.