Measuring Strength of Causal Effects with Gaussian Processes

Abstract: The objective of causal inference is to analyze the response of an effect variable to one or more cause variables. In this talk we seek to detect the presence of causal relationships among variables, characterize the nature of the causation and measure its strength. We propose a data-driven estimator of these strengths that is based on Gaussian processes. The proposed method generalizes the analysis of linear models to nonlinear relationships. Further, we show how to detect and characterize the nature of joint causal relationships. We demonstrate our approach on a number of examples and discuss its application to real-world systems.

Bio: Petar M. Djurić received bachelor's and master's degrees in electrical engineering from the University of Belgrade and a doctorate in electrical engineering from the University of Rhode Island, Kingston. He is a SUNY Distinguished Professor and currently a chair of the Department of Electrical and Computer Engineering, Stony Brook University. His research has been in the area of signal and information processing and machine learning. His recent research interests are in Bayesian machine learning with emphasis on Gaussian and deep Gaussian processes, the theory of Monte Carlo-based methods; signal and information processing over networks; and RFID and the IoT. Currently, he applies machine learning methods to problems in medicine including intrapartum fetal monitoring and brain signals. He has coauthored about 500 journal and conference papers. He has been invited to lecture at many universities in the United States and overseas. Djurić was a recipient of the IEEE Signal Processing Magazine Best Paper Award in 2007 and the EURASIP Technical Achievement Award in 2012. In 2008, he was chair of Excellence of Universidad Carlos III de Madrid-Banco de Santander. From 2008 to 2009, he was a Distinguished Lecturer of the IEEE Signal Processing Society. He has been on numerous committees of the IEEE Signal Processing Society and of many professional conferences and workshops. He was the founding editor-in-chief of the IEEE Transactions on Signal and Information Processing over Networks. He was elected a foreign member of the Serbian Academy of Engineering Sciences and is a fellow of IEEE and EURASIP.