Chandrika Kamath is a research staff member at Lawrence Livermore National Laboratory. Her current interests are in the area of scientific data mining, especially in the analysis of science data from experiments, observations, and simulations. Her multi-disciplinary expertise in algorithms includes image and video processing, feature extraction, dimension reduction, pattern recognition, machine learning, and statistical techniques.
Chandrika received her B.Tech degree in Electrical Engineering from the Indian Institute of Technology, Bombay, in 1981, followed by her M.S. in 1984 and her Ph.D. in 1986, both in Computer Science, from the University of Illinois at Urbana-Champaign. Her early career interests were in parallel numerical algorithms, especially the solution of sparse linear systems of equations. Chandrika was at Digital Equipment Corporation from 1986 through 1997, first as a Principal Software Engineer and then a Consulting Software Engineer. During this period, she developed high-performance mathematical software which became part of the Digital Extended Mathematical Library (DXML). In her later years at DEC, she was involved in the optimization and parallelization of scientific software on high-performance computers and, based on the needs of the algorithms, provided feedback to improve the compilers and system architecture. In the process, she was introduced to the mathematics underlying the Altavista search engine, prompting her to explore the emerging field of data mining.
Chandrika joined Lawrence Livermore National Laboratory in 1997, where she initially continued her work on linear solvers, investigating the Finite Element Tearing and Interconnecting method. In 1998, she started the Sapphire project to investigate how data mining techniques could be used in the analysis of massive, complex data sets common in many scientific domains. From 1998 through 2007, Chandrika was both the project lead and an individual contributor for Sapphire. In her roles, she conducted research in analysis algorithms, implemented the algorithms in software, and applied the software to practical problems at LLNL and other DOE Laboratories. The Sapphire team was awarded the 2006 R&D 100 award for their work on the scientific data mining software.
Following the successful completion of Sapphire in 2007, Chandrika has continued her work in scientific data mining. Over the years, she has been a principal investigator for a number of projects, enabling her to pursue her interests in both algorithms and applications. Some of her recent work in algorithms has focused on intelligent sampling and surrogates for simulation data, while her work in applications has addressed problems in integration of wind energy on the power grid and the determination of process parameters for metal additive manufacturing.
Chandrika holds six patents in data mining. She was involved in organizing the series of workshops on Mining Scientific Data in late 1990s and the week-long short program at the Institute for Pure and Applied Mathematics on Mathematical Challenges in Scientific Data Mining in 2002. She is active in the organization of various data mining conferences, especially the SIAM Conference on Data Mining (SDM). From 2007 through 2014, she served as the Chair of the SDM Steering Committee and was responsible for selecting and managing the team that organizes the conference each year. Her book, Scientific Data Mining:A Practical Perspective, was published by SIAM in 2009. From 2006-2009, Chandrika was the Area Editor for Applications for the Wiley journal, Statistical Analysis and Data Mining; she remains involved with the journal as one of the three Founding Editors-in-Chief. Chandrika also served as the first Chair of the SIAM Activity Group on Data Mining and Analytics (SIAG/DMA) from August 2011 through December 2013.