A Systematic Review of Challenges and Techniques of Privacy-Preserving Machine Learning

Title

A Systematic Review of Challenges and Techniques of Privacy-Preserving Machine Learning

Creator

Tiwari K.; Shukla S.; George J.P.

Description

Machine learning (ML) techniques are the backbone of Prediction and Recommendation systems, widely used across banking, medicine, and finance domains. ML techniques effectiveness depends mainly on the amount, distribution, and variety of training data that requires varied participants to contribute data. However, its challenging to combine data from multiple sources due to privacy and security concerns, competitive advantages, and data sovereignty. Therefore, ML techniques must preserve privacy when they aggregate, train, and eventually serve inferences. This survey establishes the meaning of privacy in ML, classifies current privacy threats, and describes state-of-the-art mitigation techniques named Privacy-Preserving Machine Learning (PPML) techniques. The paper compares existing PPML techniques based on relevant parameters, thereby presenting gaps in the existing literature and proposing probable future research drifts. 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Source

Lecture Notes in Networks and Systems, Vol-290, pp. 19-41.

Date

2021-01-01

Publisher

Springer Science and Business Media Deutschland GmbH

Subject

Differential privacy; Federated learning; Privacy threats; Privacy-Preserving Machine Learning; SMPC

Coverage

Tiwari K., Christ University, Bangalore, India; Shukla S., Christ University, Bangalore, India; George J.P., Christ University, Bangalore, India

Rights

Restricted Access

Relation

ISSN: 23673370; ISBN: 978-981164485-6

Format

Online

Language

English

Type

Conference paper

Identifier

https://doi.org/10.1007/978-981-16-4486-3_3

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85115095792&doi=10.1007%2f978-981-16-4486-3_3&partnerID=40&md5=6a49d3bbca944caa3da0e82a1d1a47c3