Comparison of several short-term traffic speed forecasting models

TL;DR

This paper compares various short-term traffic speed forecasting models, highlighting that traditional methods and simple adaptive models perform well without extensive training.

Contribution

It evaluates multiple forecasting models on simulated traffic data, demonstrating the effectiveness of both traditional and adaptive approaches.

Findings

Traditional techniques like regression and neural networks perform well.

Simple adaptive models without prior training also yield competitive results.

Adaptive models offer a practical alternative for real-time traffic prediction.

Abstract

The widespread adoption of smartphones in recent years has made it possible for us to collect large amounts of traffic data. Special software installed on the phones of drivers allow us to gather GPS trajectories of their vehicles on the road network. In this paper, we simulate the trajectories of multiple agents on a road network and use various models to forecast the short-term traffic speed of various links. Our results show that traditional techniques like multiple regression and artificial neural networks work well but simpler adaptive models that do not require prior training also perform comparatively well.