Regression Models for Bike Rental Prediction Based on External Factors

Wenda Li

2023

Abstract

Recently public bike systems have been more and more popular for its environmental effects and accessibility for daily use. Various researches have been conducted to improve all elements of the bike rental business, such as determining the requirements for bikes in each bike station and designing the bike recycling route. In order to develop more efficient bike systems, this paper applied two different kinds of regression models to predict the demand for rental bikes considering weather and temporal factors based on actual data in a big city. Three model frameworks are used in this paper, which are the backward stepwise selection model, ridge regression and lasso. These models show that holidays, weather situations and humidity are negative factors affecting the count while the weekday, working day, and feeling temperature are positive factors. The decision tree models are used and show better performances than linear models. The result shows that the random forest model has the best effect in predicting the demand for bikes. The study will contribute to the bike system managers to determine the redistribution of bikes depending on external factors and utilize the total amount of bikes placed in the system over a long period of time.

Paper Citation

in Harvard Style

Li W. (2023). Regression Models for Bike Rental Prediction Based on External Factors. In Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-705-4, SciTePress, pages 35-41. DOI: 10.5220/0012801900003885

in Bibtex Style

@conference{daml23,
author={Wenda Li},
title={Regression Models for Bike Rental Prediction Based on External Factors},
booktitle={Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2023},
pages={35-41},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012801900003885},
isbn={978-989-758-705-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Regression Models for Bike Rental Prediction Based on External Factors
SN - 978-989-758-705-4
AU - Li W.
PY - 2023
SP - 35
EP - 41
DO - 10.5220/0012801900003885
PB - SciTePress