Explainable AI with Information Theoretic Descriptors
Jose C. Principe
Distinguished Professor of Electrical Engineering
University of Florida
Abstract
This talk presents an overview of how information theoretic (IT) concepts and algorithms can be applied to explain the mappings learned in machine learning. The first step is to select an approach to estimate directly from data entropy and mutual information because in machine learning the pdf of the data is normally unknown. Here we will show how Renyi’s entropy and mutual information can be estimated from the eigenspectrum of the Gram matrix of kernel learning. Here we will briefly explain how IT can be used to analyze the dynamics of learning and set up proper topologies in deep learning.
Biography
Jose C. Principe (M’83-SM’90-F’00) is a Distinguished Professor of Electrical and Computer Engineering and Biomedical Engineering at the University of Florida where he teaches advanced signal processing, machine learning and artificial neural networks (ANNs) modeling. He is BellSouth Professor and the Founder and Director of the University of Florida Computational NeuroEngineering Laboratory (CNEL) www.cnel.ufl.edu . His primary area of interest is processing of time varying signals with adaptive neural models. The CNEL Lab has been studying signal and pattern recognition principles based on information theoretic criteria (entropy and mutual information).
Dr. Principe is an IEEE Fellow. He was the past Chair of the Technical Committee on Neural Networks of the IEEE Signal Processing Society, Past-President of the International Neural Network Society, and Past-Editor in Chief of the IEEE Transactions on Biomedical Engineering. He is a member of the Advisory Board of the University of Florida Brain Institute. Dr. Principe has more than 800 publications. He directed 96 Ph.D. dissertations and 65 Master theses. He wrote in 2000 an interactive electronic book entitled “Neural and Adaptive Systems” published by John Wiley and Sons and more recently co-authored several books on “Brain Machine Interface Engineering” Morgan and Claypool, “Information Theoretic Learning”, Springer, and “Kernel Adaptive Filtering”, Wiley.
Usage of Machine Learning in Brain Health
Aureli Soria-Frisch
Starlab
Biography
Dr. Aureli Soria-Frisch completed its education with a ‘Dr.-Ing.’ degree (equivalent PhD) from the Technical University Berlin in 2005. Between 1996 and 2005 he worked at the Department for Security Technologies of the Fraunhofer IPK (Berlin), where he participated in several funded research and industrial projects as research scientist and project leader. After working for 3 years at the Universitat Pompeu Fabra and part time in Starlab, he joined the company with a full engagement in 2008. He is the Director of Neuroscience Business Unit of Starlab since mid 2017. His research interest is focused on the fields: data and multi-sensory fusion, computational intelligence for data analysis, and machine learning for electrophysiological signal analysis. He has authored 20 journal papers, 9 book chapters, and over 60 conference papers. He was Project Manager of the FP7 HIVE project, and holder of 2 MJFF grants for the discovery of Machine Learning based Parkinsons’ biomarkers. He is coordinator of the H2020 FET Open project LUMINOUS, which studies human consciousness and different related technologies.