Open AccessDockless Shared Bicycle Flow Control by Using Kernel Density Estimation Based Clustering
Author(s) -
Shang-Yuan Chen,
Tzu-Tien Chen
Publication year - 2021
Publication title -
advances in technology innovation
Language(s) - English
Resource type - Journals
eISSN - 2518-2994
pISSN - 2415-0436
DOI - 10.46604/aiti.2021.6666
Subject(s) - cluster analysis , computer science , kernel density estimation , incentive , generalization , control (management) , data mining , machine learning , artificial intelligence , mathematics , statistics , economics , estimator , microeconomics , mathematical analysis
Since dockless sharing bicycles have become an indispensable means of everyday life for urban residents, how to effectively control the supply and demand balance of bikes has become an important issue. This study aims to apply Kernel Density Estimation based (KDE-based) clustering analysis and a threshold-based reverse flow incentive mechanism to encourage the users of bicycles to adjust the supply and demand actively. And it takes Shanghai Jing’an Temple and its surroundings as the research area. Its practical steps include: (1) compilation and processing of the needed data, (2) application of KDE-based clustering, partitioning, and grading, and (3) incentives calculation based on dockless shared bicycle flow control system. The study finds that the generalization function of KDE-based clustering can be used to estimate the density value at any point in the study area to support the calculation of the incentive mechanism for bicycle reverse flow.