CS&E Profile: Vipin Kumar

Data mining, bioinformatics, and high performance computing.

Ph.D. 1982, Computer Science, University of Maryland

M.E. 1979, Electronics Engineering, Phillips International Institute, Netherlands

B.E. 1977, Electronics and Communications Engineering, Indian Institute of Technology Roorkee (formerly, University of Roorkee), India

About

CS&E Department Head and William Norris Professor Vipin Kumar specializes in data mining, high performance computing and bioinformatics.

Kumar has authored more than 200 research articles, and co-edited or co-authored nine books, including widely used text books "Introduction to Parallel Computing" and "Introduction to Data Mining,'' both published by Addison Wesley. Kumar is the Editor-in-Chief of IEEE Intelligent Informatics Bulletin and the Co-Editor-in-Chief of Journal of Statistical Analysis and Data Mining. He currently serves as the chair of the steering committee of the SIAM International Conference on Data Mining, and is a member of the steering committee of the IEEE International Conference on Data Mining.

Kumar is a Fellow of the ACM, IEEE, and AAAS, and a member of SIAM. He received the 2005 IEEE Computer Society's Technical Achievement Award for contributions to the design and analysis of parallel algorithms, graph-partitioning, and data mining.

Research

My research in high-performance computing has resulted in the development of the concept of the isoefficiency metric for evaluating the scalability of parallel algorithms, as well as highly efficient parallel algorithms and state-of-the-art software for sparse matrix factorization (PSPASES), graph partitioning (METIS, ParMetis), VLSI circuit partitioning (hMetis), in addition to parallel formulations of data mining algorithms, such as those used for computing decision trees and computing associations in transaction data.

My research in data mining has focused on the development of novel techniques for analyzing massive data sets that are common in many science and engineering applications. Our group has developed novel graph-based algorithms for clustering high-dimensional data, predictive models for identifying rare events, and techniques for mining spatio-temporal data. These techniques have been used by NASA scientists for the discovery of patterns in the global climate system and teleconnections of the ocean climate to the global carbon cycle and have been incorporated in the Minnesota Intrusion Detection System (MINDS) which is a data mining based cyber security system. A new research focus in our group is on the application of data mining to the problem of protein function prediction.