Reinforcement Learning Driven Heuristic Optimization

TL;DR

This paper introduces RLHO, a reinforcement learning framework that enhances heuristic algorithms for combinatorial optimization by generating better initial solutions, leading to improved performance on problems like bin packing.

Contribution

The paper presents a novel RL-based method to improve heuristic initializations, combining reinforcement learning with traditional heuristics for better optimization results.

Findings

RLHO outperforms baselines on bin packing

RL can learn to improve heuristic initial solutions

Combining RL with heuristics enhances optimization performance

Abstract

Heuristic algorithms such as simulated annealing, Concorde, and METIS are effective and widely used approaches to find solutions to combinatorial optimization problems. However, they are limited by the high sample complexity required to reach a reasonable solution from a cold-start. In this paper, we introduce a novel framework to generate better initial solutions for heuristic algorithms using reinforcement learning (RL), named RLHO. We augment the ability of heuristic algorithms to greedily improve upon an existing initial solution generated by RL, and demonstrate novel results where RL is able to leverage the performance of heuristics as a learning signal to generate better initialization. We apply this framework to Proximal Policy Optimization (PPO) and Simulated Annealing (SA). We conduct a series of experiments on the well-known NP-complete bin packing problem, and show that the RLHO method outperforms our baselines. We show that on the bin packing problem, RL can learn to help heuristics perform even better, allowing us to combine the best parts of both approaches.