Publicação
Predicting bicycle arrivals in a Bicycle Sharing System network: A data science driven approach grounded in Zero-Inflated Regression
| Resumo: | The adoption of bicycle sharing systems (BSS) is growing in order to improve the way people move around cities, but also to stimulate the development of a more sustainable urban mobility. For the proper functioning of a BSS, it is important to have bicycles permanently available at the stations for users to start their trips, so the literature has undertaken efforts, from the perspective of the service operator, to improve the process of redistribution of bicycles and thus ensure their availability at the different stations. Since the guarantee of available bicycles cannot be assured, this work proposes to develop, from the cyclist's perspective, a proof of concept on the feasibility of informing the user about the possibility of starting a trip in a pre-defined time interval. The main contributions of this work are: (i) the ability to predict how many bicycles will arrive at a given station is a feasible improvement for BSS, (ii) the models developed through the Zero-Inflated Regression approach are a path that can be explored to improve prediction and (iii) unprecedented methodological contribution to the literature on BSS focusing on the end-user's decision power about whether or not it will soon be possible to start a trip. |
|---|---|
| Autores principais: | Leles, Weidmam Milagres |
| Assunto: | Sistema de bicicletas partilhadas SBP -- Bike share Zero-Inflated Regression Séries temporais -- Time series |
| Ano: | 2022 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | ISCTE |
| Idioma: | inglês |
| Origem: | Repositório ISCTE |
| Resumo: | The adoption of bicycle sharing systems (BSS) is growing in order to improve the way people move around cities, but also to stimulate the development of a more sustainable urban mobility. For the proper functioning of a BSS, it is important to have bicycles permanently available at the stations for users to start their trips, so the literature has undertaken efforts, from the perspective of the service operator, to improve the process of redistribution of bicycles and thus ensure their availability at the different stations. Since the guarantee of available bicycles cannot be assured, this work proposes to develop, from the cyclist's perspective, a proof of concept on the feasibility of informing the user about the possibility of starting a trip in a pre-defined time interval. The main contributions of this work are: (i) the ability to predict how many bicycles will arrive at a given station is a feasible improvement for BSS, (ii) the models developed through the Zero-Inflated Regression approach are a path that can be explored to improve prediction and (iii) unprecedented methodological contribution to the literature on BSS focusing on the end-user's decision power about whether or not it will soon be possible to start a trip. |
|---|