Adaptive Operator Selection for Many-Objective Optimization with NSGA-III

详细信息    查看全文

• 关键词：Many ; objective optimization ; Protein structure prediction ; Adaptive operator selection ; Probability matching ; Adaptive pursuit ; NSGA ; III
• 刊名：Lecture Notes in Computer Science
• 出版年：2017
• 出版时间：2017
• 年：2017
• 卷：10173
• 期：1
• 页码：267-281
• 丛书名：Evolutionary Multi-Criterion Optimization
• ISBN：978-3-319-54157-0
• 卷排序：10173

文摘

The number of objectives in real-world problems has increased in recent years and better algorithms are needed to deal efficiently with it. One possible improvement to such algorithms is the use of adaptive operator selection mechanisms in many-objective optimization algorithms. In this work, two adaptive operator selection mechanisms, Probability Matching (PM) and Adaptive Pursuit (AP), are incorporated into the NSGA-III framework to autonomously select the most suitable operator while solving a many-objective problem. Our proposed approaches, NSGA-III\(_{\text {AP}}\) and NSGA-III\(_{\text {PM}}\), are tested on benchmark instances from the DTLZ and WFG test suits and on instances of the Protein Structure Prediction Problem. Statistical tests are performed to infer the significance of the results. The preliminary results of the proposed approaches are encouraging.