Distributed Differentially Private Algorithms for Matrix and Tensor Factorization

In many signal processing and machine learning applications, datasets containing private information are held at different locations, requiring the development of distributed privacy-preserving algorithms. Tensor and matrix factorizations are key components of many processing pipelines. In the distr...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:IEEE journal of selected topics in signal processing Ročník 12; číslo 6; s. 1449 - 1464
Hlavní autori: Imtiaz, Hafiz, Sarwate, Anand D.
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: United States IEEE 01.12.2018
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Predmet:
Algorithms
Differential privacy
distributed orthogonal tensor decomposition
distributed principal component analysis
Factorization
latent variable model
Machine learning
Machine learning algorithms
Mathematical analysis
Noise
Principal component analysis
Principal components analysis
Privacy
Signal processing
Signal processing algorithms
Tensors
Differential privacy
distributed orthogonal tensor decomposition
latent variable model
distributed principal component analysis
ISSN:1932-4553, 1941-0484
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:In many signal processing and machine learning applications, datasets containing private information are held at different locations, requiring the development of distributed privacy-preserving algorithms. Tensor and matrix factorizations are key components of many processing pipelines. In the distributed setting, differentially private algorithms suffer because they introduce noise to guarantee privacy. This paper designs new and improved distributed and differentially private algorithms for two popular matrix and tensor factorization methods: principal component analysis and orthogonal tensor decomposition. The new algorithms employ a correlated noise design scheme to alleviate the effects of noise and can achieve the same noise level as the centralized scenario. Experiments on synthetic and real data illustrate the regimes in which the correlated noise allows performance matching with the centralized setting, outperforming previous methods and demonstrating that meaningful utility is possible while guaranteeing differential privacy.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:1932-4553
1941-0484
DOI:10.1109/JSTSP.2018.2877842