Abstract
We study a class of binary regularized least-squares classifiers (RLSC) for information retrieval tasks whose training involve the solution of a unique linear system of equations. Any implementation of RLSC algorithms face two major difficulties: the large size and the density of the Gram matrix. In this paper, we present a numerical investigation of an implementation based on the preconditioned conjugate gradient and introduce a novel reduced RBF kernel which is shown to improve the sparseness of the system.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Hoerl, A.E., Kennard, R.W.: Ridge regression: biased estimation for nonorthogonal problems. Technometrics 12(1), 55–67 (1970)
Joachims, T.: Text categorization with support vector machines: Learning with many relevant features. In: Nédellec, C., Rouveiol, C. (eds.) Proceedings of the European Conference on Machine Learning, pp. 137–142. Springer, Berlin (1998)
Lin, C.-J., Saigal, R.: An Incomplete Cholesky Factorization for Dense Matrices. BIT 40, 536–558 (2000)
Mukherjee, S., Niyogi, P., Rifkin, R., Poggio, T.: Statistical Learning: CVloostability is sufficient for generalization and necessary and sufficient for consistency of Empirical Risk Minimization, AI memo (2002)
Platt, J.C.: Sequential minimal optimization: A fast algorithm for training support vector machines. Technical Report MSR-TR-98-14, Microsoft Research (1998)
Poggio, T., Smale, S.: The Mathematics of Learning: Dealing with Data. Notices of the AMS (May 2003)
Rennie, J.D.M.: Using Part-of-Speech Information for Transfer in Text Classification (December 2003)
Rifkin, R.: Everything Old Is New Again: A Fresh Look at Historical Approaches in Machine Learning. PhD thesis, Massachusetts Institute of Technology (2002)
Rifkin, R., Klautau, A.: Defense of One-Vs-All Classification. Journal of Machine Learning Research 5, 101–141 (2004)
Vapnik, V.: The Nature of Statistical Learning Theory. Springer, New York (1995)
Lapack library, http://www.csit.fsu.edu/~burkardt/f_src/lapack/lapack.html
MNIST-Yann LeCun, http://yann.lecun.com/exdb/mnist/index.html
Reuters-21578 Corpus, http://www.att.research.com/~lewis/reuters21578.html
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Brahmi, A., Ech-Cherif, A. (2005). Regularized Classifiers for Information Retrieval. In: Kégl, B., Lapalme, G. (eds) Advances in Artificial Intelligence. Canadian AI 2005. Lecture Notes in Computer Science(), vol 3501. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11424918_46
Download citation
DOI: https://doi.org/10.1007/11424918_46
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-25864-3
Online ISBN: 978-3-540-31952-8
eBook Packages: Computer ScienceComputer Science (R0)