Sparse Least-Squares Methods in the Parallel Machine Learning (PML) Framework

—We describe parallel methods for solving large-scale, high-dimensional, sparse least-squares problems that arise in machine learning applications such as document classification. The basic idea is to solve a two-class response problem using a fast regression technique based on minimizing an empirical loss function, which consists of a squared-error term, and one or more regularization terms. We consider the use of Lanczos-based methods for solving these regularized least-squares problems, with the parallel implementation in the Parallel Machine Learning (PML) framework, and performance results on the IBM Blue Gene/P parallel computer. Keywords-sparse regression; classification; parallel machine learning;

• Type: Conference
• Year: 2009
• Where: ICDM
• Authors: Ramesh Natarajan, Vikas Sindhwani, Shirish Tatikonda
• Keywords: Data Mining, ICDM 2009, Least-squares Problems, Machine Learning, Parallel Machine