Scalable Fuzzy Clustering Algorithms Lawrence O. Hall Department of Computer Science and Engineering, ENB118 University of South Florida 4202 E. Fowler Ave. Tampa, Fl 33620-9951 email@example.com Abstract streams. One could also choose to cluster summarizations. Experimental data sets include Clustering is the most typical way to group several which contain tens of millions of unlabeled data. Today, there are very large examples, as well as streaming data sets. unlabeled data sets available. Many of these Results from real-world data sets show excellent data sets are too large to fit in the memory of a partitions are obtained. For tractable size data typical computer. Some of these data sets are so sets it is shown that the partitions are large that they can only be treated as data comparable to those from fuzzy c-means when streams because not all of the data can be stored it clusters all the data. in a cost-effective manner. Fuzzy clustering algorithms are known to be very useful on small Biography to medium-size data sets. This talk focuses on how to make some well understood classic Lawrence O. Hall is a Professor of Computer fuzzy clustering algorithms scale to very large Science and Engineering at University of South data sets and streaming data sets. The goal is to Florida. He received his Ph.D. in Computer be able to create a data partition that reflects the Science from the Florida State University in whole data set, but requires practical 1986 and a B.S. in Applied Mathematics from computation times. In particular, we show that the Florida Institute of Technology in 1980. He the fuzzy c-means families of algorithms can be is a fellow of the IEEE. His research interests lie scaled to provide data partitions that are very in distributed machine learning, data mining, close and potentially identical to what you pattern recognition and integrating AI into would get if you were able to cluster all the image processing. The exploitation of data. The general idea is to cluster subsets of imprecision with the use of fuzzy logic in the data and create weighted examples from the pattern recognition, AI and learning is a subsets. The weighted examples from a research theme. He has authored over 190 previous partition(s) are used with new data to publications in journals, conferences and books. create a new partition which reflects the Recent publications appear in Artificial examples currently loaded in memory and those Intelligence in Medicine, Neural Computation, partitioned previously. This process can be Pattern Recognition Letters, JAIR, Journal of repeated until all the data has been clustered. Machine Learning research, IEEE Transactions Several variations on the theme of summarizing on Systems, Man, and Cybernetics, the previous partitions with a set of weighted International Conference on Data Mining, the examples are given. Some history can be Multiple Classifier Systems Workshop, and the ignored, for example, in time changing data FUZZ-IEEE conference: (http://isl.csee.usf.edu/ailab/hall.html). He co-edited the 2001 joint North American Fuzzy Information Processing Society (NAFIPS), IFSA conference proceedings. He was the co-Program Chair of NAFIPS 2004. He received the IEEE SMC Society Outstanding contribution award in 2000. He received an Outstanding Research achievement award from the Univ. of South Florida in 2004. A past president of NAFIPS. The former vice president for membership of the SMC society. He is the President-elect of the SMC society for 2005. He is currently the Editor-In-Chief of the IEEE Transactions on Systems, Man and Cybernetics, Part B. Also, associate editor for IEEE Transactions on Fuzzy Systems, International Journal of Intelligent Data Analysis, and International Journal of Approximate Reasoning.
Pages to are hidden for
"Scalable Fuzzy Clustering Algorithms"Please download to view full document