Coupling efficiency between bike-sharing demand and land use: data envelopment analysis

Abstract

This study employs a hotspot areas-based data envelopment analysis (DEA) to investigate the coupling efficiency between bike-sharing demand and land use. Hotspot areas are used as decision-making units and re-sampling data are imposed spatial constraints. The elbow method is used to select the crucial indicators after exploring the correlation between coupling efficiency evaluation indexes using the grey correlation model. Since this coupling efficiency evaluation belongs to a ‘multi-input and single-output’ system, DEA was introduced to build a quantitative model for bike-sharing demand and land-use coupling efficiency. This paper also identifies the factors that restrict bike-sharing demand generation. More importantly, different spatial scenarios are deployed to demonstrate heterogeneous characteristics of DEA efficiency in different locations. The results of a case study in Beijing, China show that DEA efficiency generally exhibits a trend from low to high from the inside to the outside of the urban centre. The land-use factors with relatively high redundancy rates are financial insurance facilities, hotels and living facilities.