Tamara G Kolda

Bio

Tamara (Tammy) Kolda is a Distinguished Member of Technical Staff in the Informatics and Systems Assessments department at Sandia National Laboratories in Livermore, California. Her research interests include multilinear algebra and tensor decompositions, data and graph mining, optimization, nonlinear solvers, graph algorithms, cybersecurity, parallel computing and the design of scientific software. Before joining Sandia, Tammy held the Householder Postdoctoral Fellowship in Scientific Computing at Oak Ridge National Laboratory. She received her Ph.D. in applied mathematics from the University of Maryland at College Park in 1997. She has received several awards, most notably a 2003 Presidential Early Career Award for Scientists and Engineers (PECASE). She was a keynote speaker at several workshops and conferences, including the 2010 SIAM Annual Meeting and the 2007 IEEE International Conference on Data Mining (ICDM'07).

In multilinear algebra and tensor decompositions, Tammy is best known for her work on the MATLAB Tensor Toolbox and a SIAM Review article on tensor decompositions and applications. She co-authored a paper on memory-efficient Tucker (MET) tensor decompositions that resulted in the Best Paper Prize in the Theoretical/Algorithms Category at the 2008 IEEE International Conference on Data Mining (ICDM'08). In 2009, Tamara was recognized with a Sandia Award for Excellence for Laboratory Directed Research & Development for work related to tensor decompositions. In 2010, Tammy was co-organizer of the AIM Workshop on Computational Optimization for Tensor Decompositions, on the steering committee of the Conference on Tensor Decompositions and Applications, and co-organizer of the NIPS Workshop on Tensors, Kernels, and Machine Learning.

In the context of graph algorithms and models, Tammy has been active in analyzing existing models such as Stochastic Kronecker Graphs (SKG; aka R-MAT) and developed a new generative model called the Block Two-Level Erdos-Renyi (BTER) model. She is also active in applying tensor models to time-evolving graphs.

In past work, Tammy oversaw the development of HOPSPACK, a serial, multithreaded, or parallel, derivative-free optimization software framework for efficiently solving nonlinear optimization problems; this new software extends and succeeds her well-known APPSPACK software package. See her SIAM Review article on direct search methods for more information on derivative-free search. Additionally, Tammy was a contributor to the Trilinos project, a suite of numerical software packages and winner of a 2004 R&D 100 award, and co-lead on the NOX nonlinear solver C++ software package, which is part of Trilinos. Tammy has also worked in the area of graph partitioning, considering the problem of partitioning matrices for parallel computing. Tammy's thesis work considered the semi-discrete matrix decomposition (SDD) as applied to latent semantic indexing in text retrieval as well as variations on the well-known limited-memory BFGS methods in optimization.

Currently, Tammy serves on the SIAM Board of Trustees, is the Section Editor for the Software and High Performance Computing section of the SIAM Journal on Scientific Computing, and is an associate editor for SIAM Journal on Matrix Analysis. Previously, she has served on various program committees (ASONAM10, SDM10, LDMTA09, IPDPS09, SDM08, SDM06, SC02); was co-chair for the 2008 SIAM Annual Meeting; served as the Chair, Vice Chair and Secretary of the SIAM Activity Group on Computational Science and Engineering from 2009-2010, 2007-2008 and 2004-2006, respectively; served as Secretary of the SIAM Activity Group on Linear Algebra from 2001-2003; edited NA Digest from 2005-2010; served as a member of the human resources board for the American Institute of Mathematics; and was Web Editor for the Association for Women in Mathematics from 1997-2002. Tammy is also a Distinguished Member of the Association for Computing Machinery (ACM).