DisCo-DSO: Coupling Discrete and Continuous Optimization for Efficient Generative Design in Hybrid Spaces
Primary Area: general machine learning (i.e., none of the above)
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.
Keywords: Discrete Optimization, Hybrid Optimization, Deep Symbolic Optimization, Decision Trees, Reinforcement Learning, Generative Models, Interpretable Machine Learning
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
TL;DR: Novel approach for optimizing in hybrid discrete-continuous combinatorial spaces using generative models trained with reinforcement learning.
Abstract: In this paper, we consider the challenge of optimizing within hybrid discrete-continuous spaces, a problem that arises in various important applications, such as symbolic regression and decision tree learning. We propose DisCo-DSO (Discrete-Continuous Deep Symbolic Optimization), a novel approach that uses a generative model to learn a joint distribution over discrete and continuous design variables to sample new hybrid designs. In contrast to standard decoupled approaches, in which the discrete and continuous variables are optimized separately, our joint optimization approach uses fewer objective function evaluations, is robust against non-differentiable objectives, and learns from prior samples to guide the search, which leads to significant improvement in performance and efficiency. Our experiments on a diverse set of optimization tasks demonstrate that the advantages of DisCo-DSO become increasingly evident as problem complexity grows. In particular, we illustrate DisCo-DSO’s superiority over the state-of-the-art methods for interpretable reinforcement learning with decision trees.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 5807
Loading