[PAST EVENT] Colloquium Talk: Spatial-Temporal Generative Adversarial Learning
LocationMcGlothlin-Street Hall, Online on Zoom
251 Jamestown Rd
Williamsburg, VA 23185Map this location
Access & Features
- Open to the public
With the development of sensing and communication technologies, spatial-temporal big data has been widely generated and used in urban life, which helps to solve many problems related to smart cities, public safety, sustainability and business. However, it is challenging to deal with the spatial-temporal big data analytics problems (e.g., urban traffic estimation), because the data contains complex spatial-temporal dependencies, and is highly related to many other complicated factors.
In this talk, I will present an overview of my work that solves the spatial-temporal big data analytics problems in a deep generative adversarial perspective, I will also provide examples of the spatial-temporal generative adversarial learning applied in urban traffic estimation problem. My efforts along this line answer two questions: (1) how can we capture the complex spatial-temporal dependencies using generative adversarial networks (GANs)? (2) From the generative adversarial perspective, how can we learn the impacts of complex external factors? My research answers both of the questions with novel generative adversarial networks (GANs) deigned including Curb-GAN, C3-GAN, etc., which perfectly combine GAN model with unique components targeting the aforementioned problems with novel objectives, architectures and algorithms. These models can successfully solve the spatial-temporal urban data estimation problem and provide promising performance compared with other state-of-the-arts. In the end, I will conclude with a discussion of my future research directions.
Yingxue Zhang is currently a Ph.D. candidate in the Data Science Program at Worcester Polytechnic Institute, she received her BS degree in Computer Science from Shanghai Jiao Tong University in 2016, and her MS degree from Stevens Institute Technology in 2018. Her broad research interests include: (1) designing novel data mining, machine learning and AI techniques to solve spatial-temporal big data analytics problems related to smart cities, public safety, sustainability and business, and (2) human behavior analysis in autonomous driving, financial market and decision making using meta-learning, inverse reinforcement learning and imitation learning methods. Her works has been published in conferences and journals including KDD, ICDM, ACM TIST, etc.