Name Cross Validation

Intent

Motivation

Structure

<Diagram>

Discussion

Known Uses

Related Patterns

<Diagram>

References

http://www.win-vector.com/blog/2015/09/a-deeper-theory-of-testing/

http://www.slideshare.net/erogol/performance-evaluation-tutorial-46936268

https://blog.cambridgecoding.com/2016/03/24/misleading-modelling-overfitting-cross-validation-and-the-bias-variance-trade-off

https://www.cyut.edu.tw/~hchorng/downdata/1st/SS11_Variance%20Reduction.pdf

https://en.wikipedia.org/wiki/Variance_reduction

https://en.wikipedia.org/wiki/Akaike_information_criterion

https://arxiv.org/abs/1703.03167v1 Cross-validation

https://arxiv.org/abs/1704.01664 The Relative Performance of Ensemble Methods with Deep Convolutional Neural Networks for Image Classification

In this work, we investigated multiple widely used ensemble methods, including unweighted averaging, majority voting, the Bayes Optimal Classifier, and the (discrete) Super Learner, for image recognition tasks, with deep neural networks as candidate algorithms.

https://www.oreilly.com/ideas/the-preoccupation-with-test-error-in-applied-machine-learning

https://research.googleblog.com/2015/08/the-reusable-holdout-preserving.html

The reusable holdout is only one instance of a broader methodology that is, perhaps surprisingly, based on differential privacy—a notion of privacy preservation in data analysis. At its core, differential privacy is a notion of stability requiring that any single sample should not influence the outcome of the analysis significantly.

https://github.com/frankmcsherry/blog/blob/master/posts/2016-02-03.md

http://www.kdnuggets.com/2015/01/clever-methods-overfitting-avoid.html

http://andrewgelman.com/2017/07/15/what-is-overfitting-exactly/

https://arxiv.org/abs/1708.03309 Systematic Testing of Convolutional Neural Networks for Autonomous Driving

http://www.kdnuggets.com/2017/08/dataiku-predictive-model-holdout-cross-validation.html?es_p=4726597

http://memosisland.blogspot.de/2017/08/understanding-overfitting-inaccurate.html