Architecture and Knowledge-Conscious Data Mining
Amol Ghoting (The Ohio State University)
Friday, March 30th, 2007
10 am - SENSQ 5317
Refreshments at 9:30 a.m
Hosted by
Abstract
Over the past decade, our ability to gather, collect, and distribute data of all kinds has resulted in large and dynamically growing data sets at many organizations. The field of data mining, spurred by advances in data collection and storage technologies, concerns itself with the discovery of knowledge hidden in these large data sets. The knowledge discovery process is interactive in nature and thus it is imperative that we minimize response-time to a user's query. The compute- and memory-intensive nature of data mining algorithms makes this task challenging.
In this talk, I will present two design strategies to improve the performance of data mining algorithms. The first strategy, motivated by the complex nature of modern processors, strives to design architecture-conscious data mining algorithms that ensure good system utilization. I will focus on the problem domain of frequent pattern mining and show that such designs are essential to obtain performance that is commensurate with improvements in processor technology. The second strategy, motivated by the iterative nature of the knowledge discovery process, attempts to capture, maintain, and reuse repeated computation between iterations and across executions of a data mining algorithm. I will focus on the popular kMeans clustering algorithm and show that a knowledge-conscious design affords a significant performance improvement.
Biography of Speaker
Amol Ghoting is a Ph.D. candidate in the Computer Science and Engineering department at the Ohio State University. He received the M.S. degree in Computer Science from the University of Southern California in 2001 and the B.E. degree in Computer Engineering from the Victoria Jubilee Technical Institute, University of Mumbai in 2000. His research interests include data mining, database systems, high performance computing, and architecture-conscious algorithms. He is a recipient of the Best Paper Award at the 2005 International Conference on Very Large Data Bases, an Outstanding Research Award by the Ohio State University, and an IBM Ph.D. Fellowship.