Year: 2006
Type: report
Abstract: We propose a highly efficient framework for kernel multi-class models with a large and structured set of classes, and more general for penalized likelihood kernel methods. As opposed to many previous approaches which try to decompose the fitting problem into many smaller ones, we focus on a Newton optimization of the complete model, making use of model structure and linear conjugate gradients in order to approximate Newton directions. Crucially, our learning method is based entirely on matrixvector multiplication primitives with the kernel matrices and their derivatives, allowing straightforward specialization to new kernels, and directing the focus of code optimization to these primitives. Kernel parameters are learned automatically by maximizing the cross-validation log likelihood, and predictive probabilities are estimated. We demonstrate our approach on large scale text classification tasks with hierarchical structure on throusands of classes, achieving state-of-the-art results in an order of magnitude less time than previous work. We also discuss an extension to label sequence learning with kernel conditional random fields. 1
Source: MPG.PuRe (Max Planck Society)
Authors: Matthias Seeger, Olivier Chapelle
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W818938106