J. Langford, M. Seeger, and N. Megiddo. An improved predictive accuracy bound for averaging classifiers. In Proceedings of the Eighth International Conference on Machine Learning, 2001. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Discriminative, Generative and Imitative Learning - Jebara (2002) (Correct)
....the training set in the MED solution. 3. 10.3 PAC Bayes Bounds An alternative to VC dimension arguments for generalization includes PAC bounds (probably approximately correct, Valiant 1984) Recent contributions in terms of a PAC Bayesian model selection criteria by McAllester [125] and Langford [116] have given theoretical generalization arguments that directly motivate the MED approach (MED was actually developed prior to the generalization results) Essentially PAC Bayesian approaches allow the combination of a Bayesian integration of prior domain knowledge with PAC generalization ....
....the generalization results) Essentially PAC Bayesian approaches allow the combination of a Bayesian integration of prior domain knowledge with PAC generalization guarantees without forcing the PAC framework to assume the truthfulness of the prior. We loosely adapt and state the main results of [116] here but further details are available from the original work as well as [125] Effectively, the generalization guarantees are for model averaging where a stochastic model selection criterion is given in favor of a deterministic one. MED is a model averaging framework in that a distribution over ....
J. Langford, M. Seeger, and N. Megiddo. An improved predictive accuracy bound for averaging classifiers. In Proceedings of the Eighth International Conference on Machine Learning, 2001.
Data-Dependent Bounds for Bayesian Mixture Methods - Meir, Zhang (2003) (1 citation) (Correct)
....bound for the so called Gibbs algorithm, which selects a hypothesis at random from based on the posterior distribution #(h S) Essentially, this result provides a bound on the average error E h#QL(h) rather than a bound on the error of the averaged hypothesis. Later, Langford et al. [6] extended this result to a mixture of classifiers using a margin based loss function. A more general result can also be obtained using the covering number approach described in [14] Finally, Herbrich and Graepel [4] showed that under certain conditions the bounds for the Gibbs classifier can be ....
J. Langford, M. Seeger, and N. Megiddo. An improved predictive accuracy bound for averaging classifiers. In Proceeding of the Eighteenth International Conference on Machine Learning, pages 290--297, 2001.
Quantitatively Tight Sample Complexity Bounds - Langford (2002) Self-citation (Langford) (Correct)
No context found.
John Langford, Matthias Seeger, and Nimrod Megiddo, An Improved Predictive Accuracy Bound for Averaging Classi ers ICML2001.
PAC-Bayes & Margins - Langford, Shawe-Taylor Self-citation (Langford) (Correct)
No context found.
John Langford, Matthias Seeger, and Nimrod Megiddo, \An Improved Predictive Accuracy Bound for Averaging Classi ers" ICML2001.
Discriminative, Generative and Imitative Learning - Jebara (2002) (Correct)
No context found.
J. Langford, M. Seeger, and N. Megiddo. An improved predictive accuracy bound for averaging classifiers. In Proceedings of the Eighth International Conference on Machine Learning, 2001.
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