AI Benchmarks and Time Series
Speaker: Geoffrey Fox, University of Virginia
Abstract: We discuss Earthquake Prediction from the stream of past events and the use of different time series deep learning methods -- transformers and recurrent neural nets. We note that understanding the physics of the problem suggests particular observables. This problem highlights difficulties in areas with a wide dynamic range of data values. Our technology for this problem is contributed to MLCommons, and we discuss MLCommons AI benchmarks, datasets, and best practices.
Bio: Geoffrey Fox received a Ph.D. in Theoretical Physics from Cambridge University, where he was a Senior Wrangler. He is now a Professor at the Biocomplexity Institute and Initiative and in the Department of Computer Science at the University of Virginia. He previously held positions at Caltech, Syracuse University, Florida State University, and Indiana University after being a postdoc at the Institute for Advanced Study at Princeton, Lawrence Berkeley National Laboratory, and Peterhouse College Cambridge. He has supervised 75 Ph.D. students and has an h-index of 85 with over 41,000 citations. He received the High-Performance Parallel and Distributed Computing (HPDC) Achievement Award and the ACM - IEEE CS Ken Kennedy Award for Foundational contributions to parallel computing in 2019. He is a Fellow of APS (Physics) and ACM (Computing) and works on the interdisciplinary interface between computing and applications. He is currently active in the Industry consortium MLCommons/MLPerf.