A multi-head attention-based transformer model for traffic flow forecasting with a comparative analysis to recurrent neural networks

• Type: Article in International Scientific Journal
• Date: 2022-09

Authors

Selim Reza (Author) Other; Marta Campos Ferreira (Author) FEUP; José Joaquim M. Machado (Author) FEUP; João Manuel R. S. Tavares (Author) FEUP

Journal

Title: Expert Systems with Applications

Vol. 202 No. Art.117275

Pages: 1-11

ISSN: 0957-4174

Publisher: Elsevier

Indexing

ISI Web of Knowledge - 142 Citations

ISI Web of Science

Scopus - 168 Citations

Clarivate Analytics

Scientific classification

CORDIS: Technological sciences

FOS: Engineering and technology

Associated Projects

Safe Cities - Safe Cities - Inovação para Construir Cidades Seguras

Other information

Authenticus ID: P-00W-WJ1

DOI: 10.1016/j.eswa.2022.117275

Abstract (EN): Traffic flow forecasting is an essential component of an intelligent transportation system to mitigate congestion. Recurrent neural networks, particularly gated recurrent units and long short-term memory, have been the stateof-the-art traffic flow forecasting models for the last few years. However, a more sophisticated and resilient model is necessary to effectively acquire long-range correlations in the time-series data sequence under analysis. The dominant performance of transformers by overcoming the drawbacks of recurrent neural networks in natural language processing might tackle this need and lead to successful time-series forecasting. This article presents a multi-head attention based transformer model for traffic flow forecasting with a comparative analysis between a gated recurrent unit and a long-short term memory-based model on PeMS dataset in this context. The model uses 5 heads with 5 identical layers of encoder and decoder and relies on Square Subsequent Masking techniques. The results demonstrate the promising performance of the transform-based model in predicting long-term traffic flow patterns effectively after feeding it with substantial amount of data. It also demonstrates its worthiness by increasing the mean squared errors and mean absolute percentage errors by (1.25 - 47.8)% and (32.4 - 83.8)%, respectively, concerning the current baselines.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 11