A Resilient Distributed Boosting Algorithm

Yuval Filmus, Idan Mehalel, Shay Moran
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:6465-6473, 2022.

Abstract

Given a learning task where the data is distributed among several parties, communication is one of the fundamental resources which the parties would like to minimize. We present a distributed boosting algorithm which is resilient to a limited amount of noise. Our algorithm is similar to classical boosting algorithms, although it is equipped with a new component, inspired by Impagliazzo’s hard-core lemma (Impagliazzo, 1995), adding a robustness quality to the algorithm. We also complement this result by showing that resilience to any asymptotically larger noise is not achievable by a communication-efficient algorithm.

Cite this Paper


BibTeX
@InProceedings{pmlr-v162-filmus22a, title = {A Resilient Distributed Boosting Algorithm}, author = {Filmus, Yuval and Mehalel, Idan and Moran, Shay}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {6465--6473}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/filmus22a/filmus22a.pdf}, url = {https://proceedings.mlr.press/v162/filmus22a.html}, abstract = {Given a learning task where the data is distributed among several parties, communication is one of the fundamental resources which the parties would like to minimize. We present a distributed boosting algorithm which is resilient to a limited amount of noise. Our algorithm is similar to classical boosting algorithms, although it is equipped with a new component, inspired by Impagliazzo’s hard-core lemma (Impagliazzo, 1995), adding a robustness quality to the algorithm. We also complement this result by showing that resilience to any asymptotically larger noise is not achievable by a communication-efficient algorithm.} }
Endnote
%0 Conference Paper %T A Resilient Distributed Boosting Algorithm %A Yuval Filmus %A Idan Mehalel %A Shay Moran %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-filmus22a %I PMLR %P 6465--6473 %U https://proceedings.mlr.press/v162/filmus22a.html %V 162 %X Given a learning task where the data is distributed among several parties, communication is one of the fundamental resources which the parties would like to minimize. We present a distributed boosting algorithm which is resilient to a limited amount of noise. Our algorithm is similar to classical boosting algorithms, although it is equipped with a new component, inspired by Impagliazzo’s hard-core lemma (Impagliazzo, 1995), adding a robustness quality to the algorithm. We also complement this result by showing that resilience to any asymptotically larger noise is not achievable by a communication-efficient algorithm.
APA
Filmus, Y., Mehalel, I. & Moran, S.. (2022). A Resilient Distributed Boosting Algorithm. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:6465-6473 Available from https://proceedings.mlr.press/v162/filmus22a.html.

Related Material