Paper Number
ICIS2025-2369
Paper Type
Complete
Abstract
This paper presents DeepGridMCLP, a scalable two-stage method for solving the Continuous Maximal Covering Location Problem (C-MCLP). By integrating a customized Grid-Based Heuristic (GBH) for candidate location generation with a deep reinforcement learning (DRL) model for facility selection, the method addresses two major challenges in the C-MCLP: the intractability of continuous solution spaces and the inefficiency of existing MCLP solving methods. Through five experiments on synthetic and real-world datasets, including an emergency planning case study in Houston, we show how a learned policy can be used to generate scalable and high-quality solutions across diverse scenarios. DeepGridMCLP demonstrates both theoretical advances in spatial optimization within operations research and practical relevance for solving large-scale, dynamic facility placement problems such as public service planning, disaster response, and pandemic management.
Recommended Citation
Zhong, Yang and Li, Yan, "DeepGridMCLP: Scalable Facility Placement in Continuous Region with Deep Reinforcement Learning" (2025). ICIS 2025 Proceedings. 14.
https://aisel.aisnet.org/icis2025/da_bus/da_bus/14
DeepGridMCLP: Scalable Facility Placement in Continuous Region with Deep Reinforcement Learning
This paper presents DeepGridMCLP, a scalable two-stage method for solving the Continuous Maximal Covering Location Problem (C-MCLP). By integrating a customized Grid-Based Heuristic (GBH) for candidate location generation with a deep reinforcement learning (DRL) model for facility selection, the method addresses two major challenges in the C-MCLP: the intractability of continuous solution spaces and the inefficiency of existing MCLP solving methods. Through five experiments on synthetic and real-world datasets, including an emergency planning case study in Houston, we show how a learned policy can be used to generate scalable and high-quality solutions across diverse scenarios. DeepGridMCLP demonstrates both theoretical advances in spatial optimization within operations research and practical relevance for solving large-scale, dynamic facility placement problems such as public service planning, disaster response, and pandemic management.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
07-DataAnalytics