Site Selection via Learning Graph Convolutional Neural Networks: A Case Study of Singapore

Abstract

Selection of store sites is a common but challenging task in business practices. Picking the most desirable location for a future store is crucial for attracting customers and becoming profitable. The classic multi-criteria decision-making framework for store site selection oversimplifies the local characteristics that are both high dimensional and unstructured. Recent advances in deep learning enable more powerful data-driven approaches for site selection, many of which, however, overlook the interaction between different locations on the map. To better incorporate the spatial interaction patterns in understanding neighborhood characteristics and their impact on store placement, we propose to learn a graph convolutional network (GCN) for highly effective site selection tasks. Furthermore, we present a novel dataset that encompasses land use information as well as public transport networks in Singapore as a case study to benchmark site selection algorithms. It allows us to construct a geospatial GCN based on the public transport system to predict the attractiveness of different store sites within neighborhoods. We show that the proposed GCN model outperforms the competing methods that are learning from local geographical characteristics only. The proposed case study corroborates the geospatial interactions and offers new insights for solving various geographic and transport problems using graph neural networks.

Site selection and prediction of urban emergency shelter based on VGAE-RF model

“…In site selection research that emphasizes spatial relationships, researchers have analyzed the interactions among various locations on a map and developed a geospatial graph convolutional neural network (GCN) based on the public transportation system. This network predicts the attractiveness of different store locations within a community 40 . However, constructing the GCN model necessitates avoiding node overlaps to prevent data leakage, which requires the consideration of geographical locations in dataset division.…”

“…The spatial topological graph establishes connections among discrete geographic units 60,61 . Because of its ability to aggregate neighborhood information, the graph convolutional neural network (CGN) is capable of capturing the potential semantic features of nodes in a graph and has attracted the attention of many researchers 40 . Nevertheless, GCN could disrupt the overall semantics of the original graph during semi-supervised learning tasks, posing challenges in accurately capturing latent spatial patterns within the graph 62 .…”