Manifold Interpolation for Large-Scale Multiobjective Optimization via Generative Adversarial Networks

Large-scale multiobjective optimization problems (LSMOPs) are characterized as involving hundreds or even thousands of decision variables and multiple conflicting objectives. To solve LSMOPs, some algorithms designed a variety strategies to track Pareto-optimal solutions (POSs) by assuming that the distribution POSs follows low-dimensional manifold. However, traditional genetic operators for solving LSMOPs have deficiencies in dealing with manifold, which often results poor diversity, local optima, inefficient searches. In this work, generative adversarial network (GAN)-based manifold interpolation framework is proposed learn generate high-quality on thereby improving performance evolutionary algorithms. We compare approach several state-of-the-art various large-scale benchmark functions. The experimental demonstrate significant improvements been achieved LSMOPs.