交通流予測のための物理指導ニューラルネットワークに焦点を当てる。
The author proposes a physics-guided deep learning model named STDEN to bridge the gap between data-driven and physics-driven approaches in traffic flow prediction, achieving improved performance with interpretability.