Просмотр новости
Найдите то, что Вас интересует
Спутники
РН
SpaceX
Роскосмос
NASA
СпутникиРНSpaceXРоскосмосNASASpeaker: Mohammad Afzali Speaker Affiliation: UBCOctober 23, 2026ESB 4133
CanadaView All EventsAbstract: We propose a general framework for non-interactive locally differentially private (LDP) learning of structured univariate distributions that can be approximated by piece-wise polynomials. This gives the first sample-efficient LDP learners for several distribution classes, including mixtures of log-concave distributions, such as Gaussian mixture models (GMMs). These learners also provide agnostic guarantees and are computationally efficient. We also prove a lower bound for learning GMMs under LDP, improving on the standard pure-DP packing lower bounds of Bun et al. To obtain our general learning framework, we develop an efficient non-interactive LDP algorithm for synthetic data generation over large domains, which may be of independent interest. We then use it together with a deterministic adaptation of the non-private piecewise polynomial learner of Acharya et al.
Efficient Agnostic Learning of Univariate Densities Under Local Differential PrivacyEvent Topic:
Scientific Computing
Speaker:
Mohammad Afzali
Speaker Affiliation:
UBC
October 23, 2026
ESB 4133
Canada
Abstract:
We propose a general framework for non-interactive locally differentially private (LDP) learning of structured univariate distributions that can be approximated by piece-wise polynomials. This gives the first sample-efficient LDP learners for several distribution classes, including mixtures of log-concave distributions, such as Gaussian mixture models (GMMs). These learners also provide agnostic guarantees and are computationally efficient. We also prove a lower bound for learning GMMs under LDP, improving on the standard pure-DP packing lower bounds of Bun et al. To obtain our general learning framework, we develop an efficient non-interactive LDP algorithm for synthetic data generation over large domains, which may be of independent interest. We then use it together with a deterministic adaptation of the non-private piecewise polynomial learner of Acharya et al.
Event Topic:
October 23, 2026
1:00pm to 2:00pm
ESB 4133
, , CA
| # | Наименование новости | Тональность | Информативность | Дата публикации |
|---|---|---|---|---|
| 1 | Minimax density estimation in the adversarial framework under local differential privacy | 0 | 3.48 | 17-08-2026 |
| 2 | A Data-Augmented Contrastive Learning Approach to Nonparametric Density Estimation | 0 | 8.4 | 17-08-2026 |
| 3 | Nonparametric Estimation of a Factorizable Density using Diffusion Models | 0 | 8.59 | 17-08-2026 |
| 4 | Robust Representation Learning for Privacy-Preserving Machine Learning: A Multi-Objective Autoencoder Approach | 0 | 7 | 27-08-2025 |
| 5 | CPU- and GPU-Based Distributed Sampling in Dirichlet Process Mixtures for Large-Scale Analysis | 0 | 8.44 | 31-05-2026 |
| 6 | Distributed and Private Textual Data Synthesis from Embeddings | 0 | 5.6 | 01-07-2027 |
| 7 | Towards Deployable On-Device Machine Learning Systems | 0 | 13.04 | 20-07-2026 |
| 8 | Post-Quantum Stateless Steganography & MTU Bounds: Bridging P2P Censorship Resistance to Quantum Networks | 0 | 18.31 | 17-10-2026 |
| 9 | Dimensional Reduction for Sampled Priors and Application to Photometric Redshift Distributions | 0 | 5.3 | 04-08-2026 |