An interactive method for surrogate-assisted multi-objective evolutionary algorithms
| Autor: | Long Nguyen, Dinh Nguyen Duc, Kien Thai Trung |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
education.field_of_study
Computer science Process (engineering) business.industry media_common.quotation_subject Population Evolutionary algorithm Machine learning computer.software_genre Field (computer science) Support vector machine Surrogate model Kriging Quality (business) Artificial intelligence education business computer media_common |
| Zdroj: | KSE |
| DOI: | 10.1109/kse50997.2020.9287862 |
| Popis: | Solving expensive problems in the field of multi-objective optimization is an issue that researchers have been interested in and recently studied. In fact, expensive problems are quite common, which poses the need for effective methods to solve this problem. There are many methods proposed to solve expensive problems such as simulation method, decomposition method... In which, the method using a surrogate model is commonly studied and used. The algorithm of using surrogate model can use many different techniques such as RBF, PRS, Kriging, SVM, ANN... with competitive proposals. However, in order for the algorithm to be effective and applicable to problems, it is necessary to analyze and evaluate the factors that are suitable, effective, and quality. One of the factors of concern is the maintenance of the balance between the exploration and exploitation ability of the evolution process, the quality of convergence and diversity in the solution population achieved. The paper focuses on analyzing the factors affecting algorithm quality, the role of decision makers with visual reviewing to propose an interactive method to adjust algorithms to improve quality, and also meet the actual requirements of the decision maker. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|