Project Grant 2234919

Iowa State University will receive $218,212 under a two-year National Science Foundation Project Grant in the Computer and Information Science and Engineering program to develop a sub-millisecond topological feature extractor for high-rate machine learning. The university will investigate real-time topological data analysis capabilities by developing advances in mathematical, software, and hardware foundations to enable integration of topological data analysis with machine learning for modeling and forecasting complex, nonstationary time series occurring at rates measured in sub-milliseconds. The university aims to generate a field-programmable gate array design for a real-time predictor deployed at the network edge. It will customize topological data analysis approaches and produce computationally efficient machine learning representations to meet the needs of targeted modeling and forecasting tasks. The university will also identify opportunities and limitations for conducting topological data analysis in real-time across heterogeneous hardware. Expected outcomes include theoretical advances allowing integration of modern mathematical tools with hardware to demonstrate learning of nonstationary systems in real-time through generated models and forecasts.