As machine learning continues to revolutionize various scientific domains, its impact on epidemic time series forecasting has become increasingly significant. This talk explores the intersection of advanced machine learning techniques with the pressing needs of epidemic forecasting. We will address three key challenges in this field: computational efficiency, data scarcity, and the accurate modeling of spatial-temporal dynamics.
Wei Jin,
Emory University
First, I will present a solution to enable lighter, faster models to perform at the level of more complex models, crucial for timely epidemic responses. Next, I will discuss the utilization of extensive pre-training on epidemic data to improve the model’s predictive accuracy and generalization across varied epidemic scenarios. Finally, I will introduce a graph neural ODE approach for accurately modeling the continuous and dynamic spread of diseases across regions, and outline future research directions in this evolving field of machine learning for epidemic forecasting.
Wei Jin is an assistant professor of computer science at Emory University. He obtained his Ph.D. from Michigan State University in 2023. His research focuses on AI for Epidemiology, Graph Machine Learning, and Time Series Analysis, with notable accomplishments such as INNS Doctoral Dissertation Runner-up Award, AAAI New Faculty Highlights, KAUST Rising Star in AI, Snap Research Fellowship, Most Influential Papers in KDD and WWW by Paper Digest, and top finishes in three NeurIPS competitions. He has published in top-tier venues such as ICLR, KDD, ICML, NeurIPS, WWW, and AAAI, and has organized multiple tutorials and workshops at a number of these conferences. In addition, he is a passionate contributor to open-source projects and have led teams in developing several machine learning toolkits that have received downloads from more than 100 countries.