On the competitive facility location problem with a Bayesian spatial interaction model
研究企业进入新市场时如何选址和设计设施以最大化收入，提出贝叶斯空间交互模型和分层搜索算法，并在伦敦超市和酒吧数据上验证。
Abstract The competitive facility location problem arises when businesses plan to enter a new market or expand their presence. We introduce a Bayesian spatial interaction model which provides probabilistic estimates on location-specific revenues and then formulate a mathematical framework to simultaneously identify the location and design of new facilities that maximise revenue. To solve the allocation optimisation problem, we develop a hierarchical search algorithm and associated sampling techniques that explore geographic regions of varying spatial resolution. We demonstrate the approach by producing optimal facility locations and corresponding designs for two large-scale applications in the supermarket and pub sectors of Greater London.