A Comparative Study of Loss Functions: Traffic Predictions in Regular and Congestion Scenarios

TL;DR

This paper evaluates various loss functions for traffic forecasting using spatiotemporal graph neural networks, focusing on improving congestion prediction accuracy without sacrificing regular traffic speed forecasts.

Contribution

It introduces and assesses loss functions inspired by heavy tail analysis and imbalanced classification, demonstrating their effectiveness in congestion scenario forecasting.

Findings

MAE-Focal Loss outperforms others with MAE optimization.

Gumbel Loss is most effective with MSE optimization.

Enhanced congestion forecasting improves traffic management reliability.

Abstract

Spatiotemporal graph neural networks have achieved state-of-the-art performance in traffic forecasting. However, they often struggle to forecast congestion accurately due to the limitations of traditional loss functions. While accurate forecasting of regular traffic conditions is crucial, a reliable AI system must also accurately forecast congestion scenarios to maintain safe and efficient transportation. In this paper, we explore various loss functions inspired by heavy tail analysis and imbalanced classification problems to address this issue. We evaluate the efficacy of these loss functions in forecasting traffic speed, with an emphasis on congestion scenarios. Through extensive experiments on real-world traffic datasets, we discovered that when optimizing for Mean Absolute Error (MAE), the MAE-Focal Loss function stands out as the most effective. When optimizing Mean Squared Error (MSE), Gumbel Loss proves to be the superior choice. These choices effectively forecast traffic congestion events without compromising the accuracy of regular traffic speed forecasts. This research enhances deep learning models' capabilities in forecasting sudden speed changes due to congestion and underscores the need for more research in this direction. By elevating the accuracy of congestion forecasting, we advocate for AI systems that are reliable, secure, and resilient in practical traffic management scenarios.