No free lunch for cross-validation

Authors:
Huaiyu Zhu;Richard Rohwer
Affiliations:
Neural Computing Research Group, Department of Computer Science and Applied Mathematics, Aston University, Birmingham B4 7ET, UK;Neural Computing Research Group, Department of Computer Science and Applied Mathematics, Aston University, Birmingham B4 7ET, UK
Venue:
Neural Computation
Year:
1996

Citing 0
Cited 10

Note on free lunches and cross-validation

Neural Computation
On cross validation for model selection

Neural Computation
No free lunch for early stopping

Neural Computation
On different facets of regularization theory

Neural Computation
Inference for the Generalization Error

Machine Learning
No Free Lunch for Noise Prediction

Neural Computation
Antipredictable Sequences: Harder to Predict Than Random Sequences

Neural Computation
Predicting forest successional stages using multitemporal Landsat imagery with forest inventory and analysis data

International Journal of Remote Sensing
Semi-analytical method for analyzing models and model selection measures based on moment analysis

ACM Transactions on Knowledge Discovery from Data (TKDD)
Determination and the no-free-lunch paradox

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

It is known theoretically that an algorithm cannot be good for an arbitrary prior. We show that in practical terms this also applies to the technique of “cross-validation,” which has been widely regarded as defying this general rule. Numerical examples are analyzed in detail. Their implications to researches on learning algorithms are discussed.