Tong Zhang  Home/Search
Context Related
| ibm.com/dssgrp/Papers/dualcvcolt.ps Cached: PS.gz PS PDF This document uses CoBlitz to cache paper downloads. If your firewall is blocking outgoing connections to port 3125, you can use these links to download local copies.
Image Update HelpPS.gz PS PDF From: ibm.com/dssgrp/papers (more) (Enter author homepages) |
Rate this article:      (best)Comment on this article |
Abstract: In this paper, we prove a general leave-one-out style crossvalidation bound for Kernel methods. We apply this bound to some classification and regression problems, and compare the results with previously known bounds. One aspect of our analysis is that the derived expected generalization bounds reflect both approximation (bias) and learning (variance) properties of the underlying kernel methods. We are thus able to demonstrate the universality of certain learning formulations. (Update)
Similar documents based on text: More All
0.3: Relationships between Gaussian processes, Support Vector machines .. - Seeger (2000) (Correct)
0.3: Kernel Methods for Computer Vision: Theory and Applications - Camastra (Correct)
0.3: Machine Learning Strategies for Complex Tasks - Campbell, Evgeniou, Heisele.. (2000) (Correct)
Related documents from co-citation: More All
4: Princeton University Press (context) - Rockefellar, Analysis - 1970
3: The relaxation method of nding the common point of convex sets and its applicati.. (context) - Bregman - 1967
3: Statistical Learning Theory (context) - Vapnik
BibTeX entry: (Update)
Tong Zhang. A leave-one-out cross validation bound for kernel methods with applications in learning. In COLT, pages 427-443, 2001. 26 http://citeseer.ist.psu.edu/zhang01leaveoneout.html More
@inproceedings{ zhang01leaveoneout,
author = "Tong Zhang",
title = "A Leave-One-Out Cross Validation Bound for Kernel Methods with Applications in Learning",
booktitle = "14th Annual Conference on Computational Learning Theory, {COLT} 2001 and 5th {E}uropean Conference on Computational Learning Theory, {EuroCOLT} 2001, Amsterdam, The Netherlands, July 2001, Proceedings",
volume = "2111",
publisher = "Springer, Berlin",
pages = "427--443",
year = "2001",
url = "citeseer.ist.psu.edu/zhang01leaveoneout.html" }
Citations (may not include all citations):947 Statistical learning theory (context) - Vapnik - 1998
143 An Introduction to Support Vector Machines and other Kernel-.. (context) - Cristianini, Shawe-Taylor - 2000
43 Algorithmic stability and sanity-check bounds for leave-one-.. - Kearns, Ron - 1999
39 Probabilistic kernel regression models - Jaakkola, Haussler - 1999
22 Princeton University Press (context) - Rockafellar, analysis - 1970
20 Error estimates for interpolation by compactly supported rad.. - Wendland - 1998
8 Relative expected instantaneous loss bounds - uergen, Manfred - 2000
7 Algorithmic stability and generalization performance - Bousquet, Elissee - 2001
7 Approximation properties of zonal function networks using sc.. - Mhaskar, Narcowich et al. - 1999
5 Stability results for scattered-data interpolation on Euclid.. - Narcowich, Sivakumar et al. - 1998
3 Convergence of large margin separable linear classi cation - Zhang - 2001
1 A sequential approximation bound for some sample-dependent c.. (context) - Zhang - 2001
1 Relative loss bounds for temporaldi erence learning (context) - uergen, Manfred - 2000
Documents on the same site (http://www.research.ibm.com/dssgrp/papers.html): More
On the Dual Formulation of Regularized Linear Systems With Convex.. - Zhang (Correct)
A Decision-Tree-Based Symbolic Rule Induction System.. - Johnson, Oles, Zhang.. (Correct)
Text Chunking based on a Generalization of Winnow - Zhang, Damerau, Johnson (2001) (Correct)
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC