Published in Travel Behaviour and Society, 2023, 2023
We propose a multitask deep neural network framework to jointly predict mode choice and purpose in urban journeys.
Download here
Published in NeurIPS-23, 2023
We introduce the LargeST benchmark dataset, which encompasses a total number of 8,600 sensors with a 5-year time coverage and includes comprehensive metadata.
Download here
Published in NeurIPS-23, 2023
In this study, through a causal lens, we propose a novel framework called CaST to tackle temporal OoD and dynamic spatial causation in STG forecasting.
Download here
Published in SIGSPATIAL-23, 2023
This paper focuses on probabilistic STG forecasting, which is challenging due to the difficulty in modeling uncertainties and complex ST dependencies. We present the first attempt to generalize the popular denoising diffusion probabilistic models to STGs, leading to a novel non-autoregressive framework called DiffSTG, along with the first denoising network UGnet for STG in the framework.
Download here
Published in Travel Behaviour and Society, 2023, 2023
We propose a new deep learning architecture Random Effect Bayesian Neural Network (RE-BNN) to predict the travel mode choice across multiple regions and reveal regional heterogeneity.
Download here
Published in AAAI-23, 2022
In this study, we present a novel Transformer architecture termed AirFormer to collectively predict nationwide air quality in China over the next 72 hours, with an unprecedented fine spatial granularity covering thousands of locations.
Download here
Published in Sustainable Cities and Society, 2022, 2022
Taking Wuhan, China, as the study area, this paper measured the hourly accessibility to healthcare by using two methods: Gaussian Two-Step Floating Catchment Area and Weighted Average Travel Time. Then, it investigated how traffic conditions affected the equality of healthcare services among places or demographic groups at different times.
Download here