따릉이 대여소 관리를 위한 자료분석 방법 연구

초록

This study focuses on identifying the characteristics of Ttareungi users and predicting demand. It was expected that the amount of demand would change depending on the region, time, and weather of the rental station. The size of supply of bicycles would change depending on the amount of demand. So we tried to empirically understand it. Variables related to rental time, rental station, user information, and bicycle use characteristics of Seoul's Ttareungi data from October 2, 2019 to January 31, 2021 were used from the Seoul Open Data Plaza. Among the various regions in Seoul, the characteristics of users of Gangnam-gu and Gwangjin-gu by time and season were identified. There were differences in the characteristics of users by region. SNA(Social Network Analysis) is also applied for analyzing the characteristics of rental stations. Finally, Ridge, Lasso, Elastic, Lasolars, Xgboost, and randomforest were performed to predict demand by time, and the results were presented.

키워드