ar X iv : 1 50 4 . 05 99 8 v 1 [ cs . C R ] 2 2 A pr 2 01 5 Differentially Private k - Means Clustering
نویسندگان
چکیده
There are two broad approaches for differentially private data analysis. The interactive approach aims at developing customized differentially private algorithms for various data mining tasks. The non-interactive approach aims at developing differentially private algorithms that can output a synopsis of the input dataset, which can then be used to support various data mining tasks. In this paper we study the tradeoff of interactive vs. non-interactive approaches and propose a hybrid approach that combines interactive and noninteractive, using k-means clustering as an example. In the hybrid approach to differentially private k-means clustering, one first uses a non-interactive mechanism to publish a synopsis of the input dataset, then applies the standard k-means clustering algorithm to learn k cluster centroids, and finally uses an interactive approach to further improve these cluster centroids. We analyze the error behavior of both non-interactive and interactive approaches and use such analysis to decide how to allocate privacy budget between the non-interactive step and the interactive step. Results from extensive experiments support our analysis and demonstrate the effectiveness of our approach.
منابع مشابه
ar X iv : h ep - p h / 99 01 28 3 v 2 2 9 A pr 1 99 9 Diffractive charged meson pair production
We investigate the possibility to measure the nonforward gluon distribution function by means of diffractively produced π+π− and K+K− pairs in polarized lepton nucleon scattering. The resulting cross sections are small and are dominated by the gluonic contribution. We find relatively large spin asymmetries, both for π+π− and for K+K− pairs. PACS. 12.38.-t, 13.60.-r, 13.60.Le
متن کاملSpatial Opinion Dynamics and the Effects of Two Types of Mixing
Spatially-situated opinions that can be held with different degrees of conviction lead to spatio-temporal patterns such as clustering (homophily) and to polarization. Our goal is to understand how sensitive these patterns are to changes in the local nature of interactions. We introduce two different mixing mechanisms: spatial relocation, and non-local interaction (“telephoning”). We find that a...
متن کاملar X iv : h ep - p h / 95 01 29 1 v 2 4 A pr 1 99 5 EFI - 95 - 02 January 1995 hep - ph / 9501291 CHARM AND BEAUTY IN PARTICLE PHYSICS
The spectra of states containing charmed and beauty quarks, and their regularities, are reviewed.
متن کاملar X iv : c s / 01 10 03 8 v 1 [ cs . C C ] 1 8 O ct 2 00 1 Counting Is Easy †
For any fixed k, a remarkably simple single-tape Turing machine can simulate k independent counters in real time.
متن کامل