Estimation of the conditional risk in classification: The swapping method

Authors:
Jean-Jacques Daudin;Tristan Mary-Huard
Affiliations:
UMR INAPG/ENGREF/INRA MIA 518, Institut National Agronomique Paris-Grignon, 16 rue Claude Bernard, Paris Cedex 05, France;UMR INAPG/ENGREF/INRA MIA 518, Institut National Agronomique Paris-Grignon, 16 rue Claude Bernard, Paris Cedex 05, France
Venue:
Computational Statistics & Data Analysis
Year:
2008

Citing 6
Cited 2

An Experimental and Theoretical Comparison of Model SelectionMethods

Machine Learning - Special issue on the eighth annual conference on computational learning theory, (COLT '95)
Model Selection and Error Estimation

Machine Learning
Comparison of model selection for regression

Neural Computation
Classification using partial least squares with penalized logistic regression

Bioinformatics
On optimum choice of k in nearest neighbor classification

Computational Statistics & Data Analysis
Concept learning using complexity regularization

IEEE Transactions on Information Theory

Editorial: Some Recent Trends in Applied Stochastic Modeling and Multidimensional Data Analysis

Computational Statistics & Data Analysis
Measuring the prediction error. A comparison of cross-validation, bootstrap and covariance penalty methods

Computational Statistics & Data Analysis

Quantified Score

Hi-index	0.03

Visualization

Abstract

The bias of the empirical error rate in supervised classification is studied. It is shown that this bias can be understood as a covariance between the classification rule and the labeling of the training data. From this result, a new penalized criterion is proposed to perform model selection in classification. Applications of the resulting algorithm to simulated and real data are presented.