On Machine Learning “Interpretable” Representations of Dynamical Geoscientific Systems

Prof. Dr. Hoshin V. Gupta, a leading figure in hydrology and systems methods, is recognized for his work on reconciling models with data using systems theory. As a Regents Professor at the University of Arizona, he has been a KIT International Excellence Fellow at the Karlsruhe Institute of Technology for his contributions to hydrological research.
Prof. Gupta discusses how Machine Learning (ML) and Information Theory (IT) can enhance our understanding of complex geoscientific systems, such as weather and water processes. He focuses on creating models that are easy to interpret, concise, and effective in making predictions and he explains how IT helps measure information from data and models, while ML provides tools for rapid learning. This talk is suited for anyone interested in the application of modern techniques in scientific research.