Adversarial Online Learning with noise

Alon Resler, Yishay Mansour
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:5429-5437, 2019.

Abstract

We present and study models of adversarial online learning where the feedback observed by the learner is noisy, and the feedback is either full information feedback or bandit feedback. Specifically, we consider binary losses xored with the noise, which is a Bernoulli random variable. We consider both a constant noise rate and a variable noise rate. Our main results are tight regret bounds for learning with noise in the adversarial online learning model.

Cite this Paper

BibTeX
@InProceedings{pmlr-v97-resler19a, title = {Adversarial Online Learning with noise}, author = {Resler, Alon and Mansour, Yishay}, booktitle = {Proceedings of the 36th International Conference on Machine Learning}, pages = {5429--5437}, year = {2019}, editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan}, volume = {97}, series = {Proceedings of Machine Learning Research}, month = {09--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v97/resler19a/resler19a.pdf}, url = {https://proceedings.mlr.press/v97/resler19a.html}, abstract = {We present and study models of adversarial online learning where the feedback observed by the learner is noisy, and the feedback is either full information feedback or bandit feedback. Specifically, we consider binary losses xored with the noise, which is a Bernoulli random variable. We consider both a constant noise rate and a variable noise rate. Our main results are tight regret bounds for learning with noise in the adversarial online learning model.} }
Endnote
%0 Conference Paper %T Adversarial Online Learning with noise %A Alon Resler %A Yishay Mansour %B Proceedings of the 36th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Ruslan Salakhutdinov %F pmlr-v97-resler19a %I PMLR %P 5429--5437 %U https://proceedings.mlr.press/v97/resler19a.html %V 97 %X We present and study models of adversarial online learning where the feedback observed by the learner is noisy, and the feedback is either full information feedback or bandit feedback. Specifically, we consider binary losses xored with the noise, which is a Bernoulli random variable. We consider both a constant noise rate and a variable noise rate. Our main results are tight regret bounds for learning with noise in the adversarial online learning model.
APA
Resler, A. & Mansour, Y.. (2019). Adversarial Online Learning with noise. Proceedings of the 36th International Conference on Machine Learning, in Proceedings of Machine Learning Research 97:5429-5437 Available from https://proceedings.mlr.press/v97/resler19a.html.

Related Material