Differential privacy dynamics of noisy gradient descent

preview_player
Показать описание
A Google TechTalk, presented by Jiayuan Ye, National University of Singapore, at the 2021 Google Federated Learning and Analytics Workshop, Nov. 8-10, 2021.

  1. Google TechTalks
Рекомендации по теме
Differential privacy dynamics 0:10:21

Differential privacy dynamics of noisy gradient descent

How to Add 0:06:13

How to Add Noise to Achieve Differential Privacy

Differential Privacy 0:01:00

Differential Privacy Tip 1

The Skellam Mechanism 0:10:08

The Skellam Mechanism for Differentially Private Federated Learning

Cynthia Dwork: The 0:40:15

Cynthia Dwork: The Mathematics of Privacy

Understanding ε for 0:20:14

Understanding ε for Differential Privacy in Differencing Attack Scenarios

On the Convergence 0:47:24

On the Convergence of Deep Learning with Differential Privacy

Model Explanations with 0:01:27

Model Explanations with Differential Privacy

Antti Honkela: Accurate 0:44:15

Antti Honkela: Accurate privacy accounting for differentially private machine learning

Jiayuan Ye | 0:17:28

Jiayuan Ye | Differentially Private Learning Needs Hidden State (or Much Faster Convergence)

DDUO: General-Purpose Dynamic 0:10:54

DDUO: General-Purpose Dynamic Analysis for Differential Privacy

Differential Privacy 0:21:15

Differential Privacy

Differential Privacy 0:11:02

Differential Privacy

Taking advantage of 0:33:32

Taking advantage of noise to protect QML algorithms against adversaries, Nana Liu. #QRST

Privacy Amplification by 0:05:51

Privacy Amplification by Decentralization

``Differentially private inference 1:01:13

``Differentially private inference via noisy optimization'' by Marco Avella Medina

SaTML 2023 - 0:14:10

SaTML 2023 - Guy Heller - Can Stochastic Grad Langevin Dynamics Provide Differential Privacy for DL?

Solo: A Lightweight 0:07:37

Solo: A Lightweight Static Analysis for Differential Privacy

Privacy + Utility 0:15:26

Privacy + Utility Preserving Data Insights w/ State of the Art Privacy Protection| Privacy Dynamics

A Geometric View 0:54:14

A Geometric View on Private Gradient-Based Optimization

MVG Mechanism: Differential 0:22:51

MVG Mechanism: Differential Privacy under Matrix-Valued Query

Summit #11: Differential 0:29:51

Summit #11: Differential Privacy for Mobility Analytics - Alberto Hernando, Kido | Dynamics [Part 1]

Favorite Concept/Technique: Foundations 0:36:01

Favorite Concept/Technique: Foundations and Applications of Differential Privacy

SaTML 2023 - 0:59:30

SaTML 2023 - Gautam Kamath - An Introduction to Differential Privacy