Tamara Kolda, Sandia National Labs
Title: Tensor Decompositions: A Mathematical Tool for Data Analysis
Abstract: Tensors are multiway arrays, and tensor decompositions are powerful tools for data analysis and compression. In this talk, we demonstrate the wide-ranging utility of both the canonical polyadic (CP) and Tucker tensor decompositions with examples in neuroscience, chemical detection, and combustion science. The CP model is extremely useful for interpretation, as we show with an example in neuroscience. However, it can be difficult to fit to real data for a variety of reasons. We present a novel randomized method for fitting the CP decomposition to dense data that is more scalable and robust than the standard techniques. The Tucker model is useful for compression and can guarantee the accuracy of the approximation. We show that it can be used to compress massive data sets by orders of magnitude; this is done by determining the latent low-dimensional multilinear manifolds. This talk features joint work with Woody Austin (University of Texas), Casey Battaglino (Georgia Tech), Grey Ballard (Wake Forrest), Alicia Klinvex (Sandia), Hemanth Kolla (Sandia), and Alex Williams (Stanford University).
Bio: Tamara (Tammy) Kolda is a member of the Data Science and Cyber Analytics Department at Sandia National Laboratories in Livermore, CA. She earned her Ph.D. in applied mathematics from the University of Maryland at College Park. She was awarded the Householder Postdoctoral Fellowship in Scientific Computing at Oak Ridge National Lab in 1997. She joined Sandia in 1999 and was named a Distinguished Member of Technical Staff in 2010. She received a 2003 Presidential Early Career Award for Scientists and Engineers (PECASE), was named a Distinguished Scientist of the Association for Computing Machinery (ACM) in 2011 and a Fellow of the Society for Industrial and Applied Mathematics (SIAM) in 2015. She was the winner of an R&D100 award and three best paper prizes at international conferences. Her work has been cited more than 8,500 times, and her 2009 article on Tensor Decompositions and Applications is currently the #1 most-downloaded article over all SIAM journals. She has led numerous projects in computational science and data analysis on topics in multilinear algebra and tensor decompositions, graph models and algorithms, data mining, optimization, nonlinear solvers, parallel computing and the design of scientific software. She has mentored more than nineteen summer interns and six postdocs. She has given keynotes talks at a variety of meetings including the International Symposium on Mathematical Programming (ISMP), the SIAM Conference on Computational Science & Engineering (CS&E), the SIAM Annual Meeting, and the IEEE International Conference on Data Mining (ICDM). She is currently a two-time elected member of the SIAM Board of Trustees, in her second term as Section Editor for the Software and High Performance Computing Section of SISC, and in her second term as Associate Editor for SIMAX.
This seminar replaces CME 500 for this week.
Time: 4:30-5:30 p.m.
Location: Hewlett 201 *Note change in room