Keywords: heuristic search, variational autoencoders, learning to optimize, routing problems, traveling salesperson problem, vehicle routing problem, combinatorial optimization
Abstract: Methods for automatically learning to solve routing problems are rapidly improving in performance. While most of these methods excel at generating solutions quickly, they are unable to effectively utilize longer run times because they lack a sophisticated search component. We present a learning-based optimization approach that allows a guided search in the distribution of high-quality solutions for a problem instance. More precisely, our method uses a conditional variational autoencoder that learns to map points in a continuous (latent) search space to high-quality, instance-specific routing problem solutions. The learned space can then be searched by any unconstrained continuous optimization method. We show that even using a standard differential evolution search strategy our approach is able to outperform existing purely machine learning based approaches.
One-sentence Summary: We use conditional variational autoencoders to learn continuous search spaces for routing problems that can be searched with any unconstrained continuous optimizer.
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics