Fast Clustering leads to Fast SVM Training and More

Dan Boley
University of Minnesota, Twin Cities

We explore various applications that are made possible by the existence of scalable clustering algorithms. We compare deterministic and non-deterministic clusterers, and then explore how these can be used to advantage in the training of support vector machines, in handling data sets too big to fit in memory, in fast approximate latent semantic analysis, and in streaming data applications.

Audio (MP3 File, Podcast Ready) Presentation (PDF File)

Back to Workshops II: Numerical Tools and Fast Algorithms for Massive Data Mining, Search Engines and Applications