English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

New methods designed to preserve data privacy require careful scrutiny.Failure to preserve privacy is hard to detect, and yet can lead to catastrophicresults when a system implementing a ``privacy-preserving'' method is attacked.A recent work selected for an Outstanding Paper Award at ICML 2022 (Dong etal., 2022) claims that dataset condensation (DC) significantly improves dataprivacy when training machine learning models. This claim is supported bytheoretical analysis of a specific dataset condensation technique and anempirical evaluation of resistance to some existing membership inferenceattacks.  In this note we examine the claims in the work of Dong et al. (2022) anddescribe major flaws in the empirical evaluation of the method and itstheoretical analysis. These flaws imply that their work does not providestatistically significant evidence that DC improves the privacy of training MLmodels over a naive baseline. Moreover, previously published results show thatDP-SGD, the standard approach to privacy preserving ML, simultaneously givesbetter accuracy and achieves a (provably) lower membership attack success rate.

No Free Lunch in "Privacy for Free: How does Dataset Condensation Help Privacy"