Many-Objective Estimation of Distribution Optimization Algorithm Based on WGAN-GP
This work addresses the challenge of solving many-objective optimization problems (MaOPs) for researchers and practitioners in optimization and evolutionary algorithms, representing an incremental improvement by integrating WGAN-GP into an existing EDA framework.
The paper tackles the performance decline of Estimation of Distribution Algorithms (EDA) in many-objective optimization problems (MaOPs) by replacing crossover and mutation with Wasserstein Generative Adversarial Networks-Gradient Penalty (WGAN-GP) within the Reference Vector Guided Evolutionary Algorithm (RVEA) framework, resulting in improved convergence and diversity as demonstrated through comparisons with RM-MEDA, MOPSO, and NSGA-II on DTLZ and LSMOP test suites with up to 15 objectives.
Estimation of distribution algorithms (EDA) are stochastic optimization algorithms. EDA establishes a probability model to describe the distribution of solution from the perspective of population macroscopically by statistical learning method, and then randomly samples the probability model to generate a new population. EDA can better solve multi-objective optimal problems (MOPs). However, the performance of EDA decreases in solving many-objective optimal problems (MaOPs), which contains more than three objectives. Reference Vector Guided Evolutionary Algorithm (RVEA), based on the EDA framework, can better solve MaOPs. In our paper, we use the framework of RVEA. However, we generate the new population by Wasserstein Generative Adversarial Networks-Gradient Penalty (WGAN-GP) instead of using crossover and mutation. WGAN-GP have advantages of fast convergence, good stability and high sample quality. WGAN-GP learn the mapping relationship from standard normal distribution to given data set distribution based on a given data set subject to the same distribution. It can quickly generate populations with high diversity and good convergence. To measure the performance, RM-MEDA, MOPSO and NSGA-II are selected to perform comparison experiments over DTLZ and LSMOP test suites with 3-, 5-, 8-, 10- and 15-objective.