Graph Reinforcement Learning for Operator Selection in the ALNS Metaheuristic
This work addresses a long-standing open question in ALNS research, offering a practical improvement for combinatorial optimization practitioners, though it appears incremental as it builds on existing ALNS and reinforcement learning methods.
The paper tackled the problem of efficiently selecting operators within the ALNS metaheuristic for combinatorial optimization by formulating it as a Markov Decision Process and using Deep Reinforcement Learning with Graph Neural Networks. The result showed better performance than the classic adaptive layer, with potential time and labor savings in handcrafting operator portfolios.
ALNS is a popular metaheuristic with renowned efficiency in solving combinatorial optimisation problems. However, despite 16 years of intensive research into ALNS, whether the embedded adaptive layer can efficiently select operators to improve the incumbent remains an open question. In this work, we formulate the choice of operators as a Markov Decision Process, and propose a practical approach based on Deep Reinforcement Learning and Graph Neural Networks. The results show that our proposed method achieves better performance than the classic ALNS adaptive layer due to the choice of operator being conditioned on the current solution. We also discuss important considerations such as the size of the operator portfolio and the impact of the choice of operator scales. Notably, our approach can also save significant time and labour costs for handcrafting problem-specific operator portfolios.