Machine Learning for Smart Cities: LSTM Model-Based Taxi OD Demand Forecasting in New York

Ziyan Chen

2024

Abstract

This study delves into the realm of advanced machine learning techniques, with a particular focus on employing the Long Short-term Memory Network (LSTM) model for forecasting Taxi Origin-Destination (OD) demand in New York City. In the quest for the most accurate predictive model, this paper conducted a comparative analysis between the Decision Tree (DT), Random Forest (RF), and the aforementioned LSTM model. The findings of this study reveal that the LSTM model outperforms its counterparts in both prediction accuracy and generalization capability. The model's coefficient of determination (R²) stands at an impressive 0.9657, signifying that it captures a substantial 96.57% of the variance within the dataset. Through model optimization, this study has further minimized the error index, highlighting the sensitivity of the model to its configuration and the potential for enhanced performance. As looking towards the horizon, future research endeavors will concentrate on overcoming current limitations and bolstering the robustness and applicability of the LSTM model. The further study plans to extend its application to various urban settings and integrate real-time data streams to augment its predictive prowess. Additionally, examining the model's efficacy in a multi-modal traffic context and exploring the synthesis of LSTM with other machine learning algorithms to forge hybrid models could lead to the development of more sophisticated and precise demand forecasting tools. These advancements will be instrumental in facilitating smarter urban transport planning and management, thereby revolutionizing the way approaching Taxi OD demand forecasting in the era of machine learning.

Paper Citation

in Harvard Style

Chen Z. (2024). Machine Learning for Smart Cities: LSTM Model-Based Taxi OD Demand Forecasting in New York. In Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI; ISBN 978-989-758-726-9, SciTePress, pages 27-32. DOI: 10.5220/0013205400004568

in Bibtex Style

@conference{ecai24,
author={Ziyan Chen},
title={Machine Learning for Smart Cities: LSTM Model-Based Taxi OD Demand Forecasting in New York},
booktitle={Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI},
year={2024},
pages={27-32},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013205400004568},
isbn={978-989-758-726-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI
TI - Machine Learning for Smart Cities: LSTM Model-Based Taxi OD Demand Forecasting in New York
SN - 978-989-758-726-9
AU - Chen Z.
PY - 2024
SP - 27
EP - 32
DO - 10.5220/0013205400004568
PB - SciTePress