Traffic Prediction Using LSTM, RF and XGBoost

Ka Nam Lam

2024

Abstract

Traffic congestion is one of the most challenging and lasting problems that causes many government concerns. It would lead to many problems, such as economic losses, fuel consumption, environmental costs, and so on. An efficient traffic system can significantly reduce congestion, which can bring many beneficial impacts on daily life. Accurate traffic flow prediction is crucial for effective traffic management. This study uses three machine learning models: Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Long Short-Term Memory (LSTM) to predict vehicle counts at four different junctions of a city. Each of these models was evaluated based on key metrics – Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R-squared (𝑅2). The outcomes showed that XGBoost performed the best among the examined models in terms of precision and computational efficiency. This paper also discusses the limitations of the models and future implications, which can be helpful in better managing transportation systems.

Paper Citation

in Harvard Style

Lam K. (2024). Traffic Prediction Using LSTM, RF and XGBoost. In Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-754-2, SciTePress, pages 267-274. DOI: 10.5220/0013515600004619

in Bibtex Style

@conference{daml24,
author={Ka Nam Lam},
title={Traffic Prediction Using LSTM, RF and XGBoost},
booktitle={Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2024},
pages={267-274},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013515600004619},
isbn={978-989-758-754-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Traffic Prediction Using LSTM, RF and XGBoost
SN - 978-989-758-754-2
AU - Lam K.
PY - 2024
SP - 267
EP - 274
DO - 10.5220/0013515600004619
PB - SciTePress