loading page

Dynamic Multi-objective Estimation of Distribution Algorithm Based on Domain Adaptation and Nonparametric Estimation
  • Min JIANG
Min JIANG

Corresponding Author:[email protected]

Author Profile

Abstract

Although dynamic optimization and multi-objective optimization have made considerable progress separately, solving dynamic multi-objective optimization problems remains a challenging topic since its objective functions change during optimization. In this paper, we propose a Domain Adaptation and Nonparametric Estimation based Estimation of Distribution Algorithm (EDA), called DANE-EDA, to solve dynamic multi-objective optimization problem. The feature of the proposed algorithm is that the importance sampling, nonparametric density estimation, probabilistic prediction mechanism and domain adaptation technique are unified under one framework, so that it can take the advantages of the Monte-Carlo method and transfer learning method. This kind of combination will help the proposed algorithm to keep the exploration-exploitation tradeoff from time and spatial perspectives as well as remedies the disadvantage, such as the diversity of the solutions decreased, caused by the machine learning technology. After proving the convergence and computational complexity of the DANE-EDA algorithm, we compared the proposed method with nine EDAs or dyamic multiobjective optimization algorithms on twelve different test instances. The experimental results proved the effectiveness of the proposed method.