SF Bay ACM Data Mining SIG: Machine Learning on Multi-Core November 8, 2006

We are at the beginning of the multicore era. Computers will have increasingly many cores (processors), but there is still no good programming framework for these architectures, and thus no simple and unified way for machine learning to take advantage of the potential speedup afforded by multicore. We devise a certain "summation form," which allows conforming algorithms to be easily parallelized on multicore computers. We adapt Google's map-reduce paradigm to demonstrate this parallel speedup technique on a wide variety of machine learning and computer vision algorithms. We show that this programming framework is pragmatic and easy to learn for taking advantage of multicore parallelism. Our experimental results show basically linear speedup with an increasing number of processors.

Official Website: http://www.sfbayacm.org/events/2006-11-15.php