Optimized Training of Support Vector Machines on the Cell Processor

Support Vector Machines (SVMs) are a widely used technique for classification, clustering and data analysis. While efficient algorithms for training SVMs are available, dealing with large datasets makes training and classification a computationally challenging problem. In this paper we exploit modern processor architectures to improve the training speed of LIBSVM, a well known software tool which implements the Sequential Minimal Optimization algorithm. We describe LIBSVMCBE , an optimized version of LIBSVM which takes advantage of the peculiar architecture of the Cell Processor. We assess the performance of LIBSVMCBE on real-world training problems, and we show how this optimization is particularly effective on large, dense datasets. 1. Department of Computer Science, University of Bologna, Mura Anteo Zamboni 7, I-40127 Bologna, Italy, Email: [email protected]