An interpretable machine learning framework to understand bikeshare demand before and during the COVID-19 pandemic in New York City

Majbah Uddin; Ho Ling Hwang; Md Sami Hasnine

doi:10.1080/03081060.2023.2201280

An interpretable machine learning framework to understand bikeshare demand before and during the COVID-19 pandemic in New York City

Majbah Uddin, Ho Ling Hwang, Md Sami Hasnine

Transportation Analytics & Decision Scie

Research output: Contribution to journal › Article › peer-review

3 Scopus citations

Abstract

In recent years, bikesharing systems have become increasingly popular as affordable and sustainable micromobility solutions. Advanced mathematical models such as machine learning are required to generate good forecasts for bikeshare demand. To this end, this study proposes a machine learning modeling framework to estimate hourly demand in a large-scale bikesharing system. Two Extreme Gradient Boosting models were developed: one using data from before the COVID-19 pandemic (March 2019 to February 2020) and the other using data from during the pandemic (March 2020 to February 2021). Furthermore, a model interpretation framework based on SHapley Additive exPlanations was implemented. Based on the relative importance of the explanatory variables considered in this study, share of female users and hour of day were the two most important explanatory variables in both models. However, the month variable had higher importance in the pandemic model than in the pre-pandemic model.

Original language	English
Pages (from-to)	482-498
Number of pages	17
Journal	Transportation Planning and Technology
Volume	46
Issue number	4
DOIs	https://doi.org/10.1080/03081060.2023.2201280
State	Published - 2023

Funding

This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan ( http://energy.gov/downloads/doe-public-access-plan ). This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).

Keywords

Bikeshare
Citi Bike
New York City
SHAP
machine learning

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1080/03081060.2023.2201280

Cite this

@article{a4e9cadcbfd54bc89c64349e84b0c603,

title = "An interpretable machine learning framework to understand bikeshare demand before and during the COVID-19 pandemic in New York City",

abstract = "In recent years, bikesharing systems have become increasingly popular as affordable and sustainable micromobility solutions. Advanced mathematical models such as machine learning are required to generate good forecasts for bikeshare demand. To this end, this study proposes a machine learning modeling framework to estimate hourly demand in a large-scale bikesharing system. Two Extreme Gradient Boosting models were developed: one using data from before the COVID-19 pandemic (March 2019 to February 2020) and the other using data from during the pandemic (March 2020 to February 2021). Furthermore, a model interpretation framework based on SHapley Additive exPlanations was implemented. Based on the relative importance of the explanatory variables considered in this study, share of female users and hour of day were the two most important explanatory variables in both models. However, the month variable had higher importance in the pandemic model than in the pre-pandemic model.",

keywords = "Bikeshare, Citi Bike, New York City, SHAP, machine learning",

author = "Majbah Uddin and Hwang, {Ho Ling} and Hasnine, {Md Sami}",

note = "Publisher Copyright: {\textcopyright} 2023 Informa UK Limited, trading as Taylor & Francis Group.",

year = "2023",

doi = "10.1080/03081060.2023.2201280",

language = "English",

volume = "46",

pages = "482--498",

journal = "Transportation Planning and Technology",

issn = "0308-1060",

publisher = "Taylor and Francis Group",

number = "4",

}

TY - JOUR

T1 - An interpretable machine learning framework to understand bikeshare demand before and during the COVID-19 pandemic in New York City

AU - Uddin, Majbah

AU - Hwang, Ho Ling

AU - Hasnine, Md Sami

PY - 2023

Y1 - 2023

N2 - In recent years, bikesharing systems have become increasingly popular as affordable and sustainable micromobility solutions. Advanced mathematical models such as machine learning are required to generate good forecasts for bikeshare demand. To this end, this study proposes a machine learning modeling framework to estimate hourly demand in a large-scale bikesharing system. Two Extreme Gradient Boosting models were developed: one using data from before the COVID-19 pandemic (March 2019 to February 2020) and the other using data from during the pandemic (March 2020 to February 2021). Furthermore, a model interpretation framework based on SHapley Additive exPlanations was implemented. Based on the relative importance of the explanatory variables considered in this study, share of female users and hour of day were the two most important explanatory variables in both models. However, the month variable had higher importance in the pandemic model than in the pre-pandemic model.

AB - In recent years, bikesharing systems have become increasingly popular as affordable and sustainable micromobility solutions. Advanced mathematical models such as machine learning are required to generate good forecasts for bikeshare demand. To this end, this study proposes a machine learning modeling framework to estimate hourly demand in a large-scale bikesharing system. Two Extreme Gradient Boosting models were developed: one using data from before the COVID-19 pandemic (March 2019 to February 2020) and the other using data from during the pandemic (March 2020 to February 2021). Furthermore, a model interpretation framework based on SHapley Additive exPlanations was implemented. Based on the relative importance of the explanatory variables considered in this study, share of female users and hour of day were the two most important explanatory variables in both models. However, the month variable had higher importance in the pandemic model than in the pre-pandemic model.

KW - Bikeshare

KW - Citi Bike

KW - New York City

KW - SHAP

KW - machine learning

UR - http://www.scopus.com/inward/record.url?scp=85153200911&partnerID=8YFLogxK

U2 - 10.1080/03081060.2023.2201280

DO - 10.1080/03081060.2023.2201280

M3 - Article

AN - SCOPUS:85153200911

SN - 0308-1060

VL - 46

SP - 482

EP - 498

JO - Transportation Planning and Technology

JF - Transportation Planning and Technology

IS - 4

ER -

An interpretable machine learning framework to understand bikeshare demand before and during the COVID-19 pandemic in New York City

Abstract

Funding

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this