VColRL: Learn to Solve the Vertex Coloring Problem Using Reinforcement Learning

Abhinav Anand, Subrahmanya Swamy Peruru, Amitangshu Pal

Keywords: Vertex Coloring Problem, Markov Decision Process, Reinforcement Learning, Graph Neural Networks, Combinatorial Optimization

TL;DR: VColRL is a reinforcement learning framework that solves the vertex coloring problem.

Abstract: We present VColRL, a reinforcement learning framework designed to solve the vertex coloring problem (VCP), where the objective is to assign colors to the vertices of a graph with the minimum number of colors, such that no two adjacent vertices share the same color. The framework is built on a novel Markov Decision Process (MDP) configuration to effectively capture the dynamics of the VCP, developed after evaluating various MDP configurations. Our experimental results demonstrate that VColRL achieves competitive performance in terms of using fewer colors as compared to advanced mathematical solvers and other metaheuristic approaches while being significantly faster. Additionally, our results show that VColRL generalizes well across different types of graphs.

Primary Area: reinforcement learning

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.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

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: 9921