Explainability techniques applied to road traffic forecasting using Graph Neural Network models

Please use this identifier to cite or link to this item: http://hdl.handle.net/10045/135268
Información del item - Informació de l'item - Item information
Title: Explainability techniques applied to road traffic forecasting using Graph Neural Network models
Authors: García-Sigüenza, Javier | Llorens Largo, Faraón | Tortosa, Leandro | Vicent, Jose F.
Research Group/s: Análisis y Visualización de Datos en Redes (ANVIDA) | Grupo de Investigación en Tecnologías Inteligentes para el Aprendizaje (Smart Learning)
Center, Department or Service: Universidad de Alicante. Departamento de Ciencia de la Computación e Inteligencia Artificial
Keywords: Graph neural networks | Deep learning | Data analysis | Explainability | Traffic flow
Issue Date: 16-Jun-2023
Publisher: Elsevier
Citation: Information Sciences. 2023, 645: 119320. https://doi.org/10.1016/j.ins.2023.119320
Abstract: In recent years, several new Artificial Intelligence methods have been developed to make models more explainable and interpretable. The techniques essentially deal with the implementation of transparency and traceability of black box machine learning methods. Black box refers to the inability to explain why the model turns the input into the output, which may be problematic in some fields. To overcome this problem, our approach provides a comprehensive combination of predictive and explainability techniques. Firstly, we compared statistical regression, classic machine learning and deep learning models, reaching the conclusion that models based on deep learning exhibit greater accuracy. Of the great variety of deep learning models, the best predictive model in spatio-temporal traffic datasets was found to be the Adaptive Graph Convolutional Recurrent Network. Regarding the explainability technique, GraphMask shows a notably higher fidelity metric than other methods. The integration of both techniques was tested by means of experimental results, concluding that our approach improves deep learning model accuracy, making such models more transparent and interpretable. It allows us to discard up to 95% of the nodes used, facilitating an analysis of its behavior and thus improving the understanding of the model.
URI: http://hdl.handle.net/10045/135268
ISSN: 0020-0255 (Print) | 1872-6291 (Online)
DOI: 10.1016/j.ins.2023.119320
Language: eng
Type: info:eu-repo/semantics/article
Rights: © 2023 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer Review: si
Publisher version: https://doi.org/10.1016/j.ins.2023.119320
Appears in Collections:INV - Smart Learning - Artículos de Revistas
INV - ANVIDA - Artículos de Revistas

Files in This Item:
Files in This Item:
File Description SizeFormat 
ThumbnailGarcia-Siguenza_etal_2023_InformatSci.pdf6,25 MBAdobe PDFOpen Preview

This item is licensed under a Creative Commons License Creative Commons