The Bias-Variance Tradeoff and the Randomized GACV

Abstract

We propose a new in-sample cross validation based method (randomized GACV) for choosing smoothing or bandwidth parameters that govern the bias-variance or fit-complexity tradeoff in 'soft' classification. Soft clas(cid:173) sification refers to a learning procedure which estimates the probability that an example with a given attribute vector is in class 1 vs class O. The target for optimizing the the tradeoff is the Kullback-Liebler distance between the estimated probability distribution and the 'true' probabil(cid:173) ity distribution, representing knowledge of an infinite population. The method uses a randomized estimate of the trace of a Hessian and mimics cross validation at the cost of a single relearning with perturbed outcome data.

Cite

Text

Wahba et al. "The Bias-Variance Tradeoff and the Randomized GACV." Neural Information Processing Systems, 1998.

Markdown

[Wahba et al. "The Bias-Variance Tradeoff and the Randomized GACV." Neural Information Processing Systems, 1998.](https://mlanthology.org/neurips/1998/wahba1998neurips-biasvariance/)

BibTeX

@inproceedings{wahba1998neurips-biasvariance,
  title     = {{The Bias-Variance Tradeoff and the Randomized GACV}},
  author    = {Wahba, Grace and Lin, Xiwu and Gao, Fangyu and Xiang, Dong and Klein, Ronald and Klein, Barbara},
  booktitle = {Neural Information Processing Systems},
  year      = {1998},
  pages     = {620-626},
  url       = {https://mlanthology.org/neurips/1998/wahba1998neurips-biasvariance/}
}