Short-Term Traffic Flow Forecasting A wide and deep approach with periodic feature selection
Published in TechRxiv, 2022
This paper presents a novel approach to traffic flow forecasting that utilizes a hybrid wide and deep learning architecture. This model integrates both spatial-temporal and periodic features, aiming to enhance the predictive accuracy of traffic forecasting systems. The wide component of the model focuses on capturing periodic features, particularly the weekly patterns, shown to be most impactful, while the deep component, built on a conv-LSTM architecture, extracts spatial-temporal features. The effectiveness of this model is demonstrated through extensive experiments that compare its performance against traditional models, showing notable improvements in forecasting accuracy.
Recommended citation: Martin Esugo, Qian Lu, Olivier Haas. "Short-Term Traffic Flow Forecasting A wide and deep approach with periodic feature selection." TechRxiv. May 17, 2022 http://komehz.github.io/files/2022-05-17-paper-002.pdf