Seminar by Dr. Peter Song, University of Michigan

Abstract

In this talk I will introduce an incremental learning algorithm to analyze streaming datasets with correlated outcomes such as longitudinal data and clustered data. We develop a renewable QIF (RenewQIF) method within a paradigm of renewable estimation and incremental inference, in which statistical results are recursively renewed with a current data batch and summary statistics of historical data batches, but with no use of any historical subject-level raw data. We compare our renewable estimation method with both offline QIF method and offline generalized estimating equations (GEE) approach that process the entire cumulative subject-level data all together. We show theoretically and numerically that the RenewQIF enjoys statistical and computational efficiency. In addition, we propose an approach to diagnose the homogeneity assumption of regression coefficients via a sequential goodness-of-fit test as a screening procedure on occurrences of potential abnormal data batches. We implement the proposed methodology by expanding the existing Spark’s Lambda architecture for the operation of statistical inference and data quality monitoring. We illustrate the proposed methodology by extensive simulation studies and an analysis of streaming car crash datasets from the National Automotive Sampling System-Crashworthiness Data System (NASS CDS).

Bio

Dr. Song is Professor of Biostatistics at the Department of Biostatistics, School of Public Health in the University of Michigan, Ann Arbor. He received his PhD in Statistics from the University of British Columbia, Vancouver, Canada in 1996. He has published over 190 peer-reviewed papers and graduated 22 PhD students. Dr. Song's current research interests include data integration, distributed inference, high-dimensional data analysis, longitudinal data analysis, mediation analysis, spatiotemporal modeling, and smart health. He collaborates extensively with researchers from nutritional sciences, environmental health sciences, chronic diseases, and nephrology. He is IMS Fellow, ASA Fellow and Elected Member of the International Statistical Institute. Dr. Song now serves as Associate Editor of the Journal of American Statistical Association, the Annals of Applied Statistics, and the Journal of Multivariate Analysis.