Abstract

We consider the problem of PAC-learning from distributed data and analyze fundamental communication complexity questions involved. We provide general upper and lower bounds on the amount of communication needed to learn well, showing that in addition to VC-dimension and covering number, quantities such as the teaching-dimension and mistake-bound of a class play an important role. We also present tight results for a number of common concept classes including conjunctions, parity functions, and decision lists. For linear separators, we show that for non-concentrated distributions, we can use a version of the Perceptron algorithm to learn with much less communication than the number of updates given by the usual margin bound. We also show how boosting can be performed in a generic manner in the distributed setting to achieve communication with only logarithmic dependence on 1/epsilon for any concept class, and demonstrate how recent work on agnostic learning from class-conditional queries can be used to achieve low communication in agnostic settings as well. We additionally present an analysis of privacy, considering both differential privacy and a notion of distributional privacy that is especially appealing in this context.

Keywords

Computer scienceCommunication complexityLogarithmUpper and lower boundsConcept classDifferential privacyTheoretical computer scienceMistakeVC dimensionClass (philosophy)Dimension (graph theory)Boosting (machine learning)Margin (machine learning)Context (archaeology)Artificial intelligenceMachine learningMathematicsAlgorithmCombinatorics

Affiliated Institutions

Related Publications

Publication Info

Year
2012
Type
preprint
Volume
23
Citations
113
Access
Closed

External Links

Social Impact

Social media, news, blog, policy document mentions

Citation Metrics

113
OpenAlex

Cite This

Maria-Florina Balcan, Avrim Blum, Shai Fine et al. (2012). Distributed Learning, Communication Complexity and Privacy. arXiv (Cornell University) , 23 . https://doi.org/10.48550/arxiv.1204.3514

Identifiers

DOI
10.48550/arxiv.1204.3514