Projects : fast_learning

We explore algorithms for implementing large scale learning algorithms as parallel computation. We are currently developing parallelization approaches for increasing the ability of SVM (Support Vector Machines) to solve large-scale problems. As target systems, we consider shared memory processors, clusters of processors, vector processors, and SIMD (Single Instruction Multiple Data) processors. On a given system the speed of an SVM is limited by the compute performance of the processor as well as by the size of the memory. Efficient parallelizations have to overcome both of these limitations while not getting bogged down in communication overhead.

Online learning gives the promise of dealing with very large datasets because data can be streamed off of a disk or another source, and the entire dataset does not have to be held in memory. We investigate fast online algorithms that achieve good generalization ability after only one pass of the data.

Transduction and semi-supervised learning methods can help improve generalization ability in learning problems through the use of the test labels, or unlabeled data, during learning. However, many algorithms are unfeasibly slow. We investigate how to make large scale algorithms in this domain.