Proactive DP: A Multple Target Optimization Framework for DP-SGD

• URL: http://arxiv.org/abs/2102.09030v9
• Date: Fri, 24 Nov 2023 11:25:40 GMT
• Title: Proactive DP: A Multple Target Optimization Framework for DP-SGD
• Authors: Marten van Dijk, Nhuong V. Nguyen, Toan N. Nguyen, Lam M. Nguyen and Phuong Ha Nguyen
• Abstract summary: We introduce a multiple target optimization framework for DP-SGD referred to as pro-active DP. The pro-active DP scheme allows one to it a-priori select parameters of DP-SGD based on a fixed privacy budget. We present a closed-form $(epsilon,delta)$-DP guarantee that connects all parameters in the DP-SGD setup.
• Score: 22.663029794467768
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce a multiple target optimization framework for DP-SGD referred to as pro-active DP. In contrast to traditional DP accountants, which are used to track the expenditure of privacy budgets, the pro-active DP scheme allows one to {\it a-priori} select parameters of DP-SGD based on a fixed privacy budget (in terms of $\epsilon$ and $\delta$) in such a way to optimize the anticipated utility (test accuracy) the most. To achieve this objective, we first propose significant improvements to the moment account method, presenting a closed-form $(\epsilon,\delta)$-DP guarantee that connects all parameters in the DP-SGD setup. Generally, DP-SGD is $(\epsilon\leq 1/2,\delta=1/N)$-DP if $\sigma=\sqrt{2(\epsilon +\ln(1/\delta))/\epsilon}$ with $T$ at least $\approx 2k^2/\epsilon$ and $(2/e)^2k^2-1/2\geq \ln(N)$, where $T$ is the total number of rounds, and $K=kN$ is the total number of gradient computations where $k$ measures $K$ in number of epochs of size $N$ of the local data set. We prove that our expression is close to tight in that if $T$ is more than a constant factor $\approx 4$ smaller than the lower bound $\approx 2k^2/\epsilon$, then the $(\epsilon,\delta)$-DP guarantee is violated. Our enhanced DP theory allows us to create a utility graph and DP calculator. These tools link privacy and utility objectives and search for optimal experiment setups, efficiently taking into account both accuracy and privacy objectives, as well as implementation goals. We furnish a comprehensive implementation flow of our proactive DP, with rigorous experiments to showcase the proof-of-concept.

Related papers

• Optimal Rates for $O(1)$-Smooth DP-SCO with a Single Epoch and Large Batches [12.184984662899868]
  We revisit the correlated convex optimization (SCO) problem. We propose an algorithm that achieves the optimal rate for DP-SCO (up to polylog factors) in a single epoch.
  arXiv  Detail & Related papers  (2024-06-04T18:59:42Z)
• Private Mean Estimation with Person-Level Differential Privacy [6.621676316292624]
  We study person-level differentially private mean estimation in the case where each person holds multiple samples. We give computationally efficient algorithms under approximate-DP and computationally inefficient algorithms under pure DP, and our nearly matching lower bounds hold for the most permissive case of approximate DP.
  arXiv  Detail & Related papers  (2024-05-30T18:20:35Z)
• PrivSGP-VR: Differentially Private Variance-Reduced Stochastic Gradient Push with Tight Utility Bounds [9.47030623916154]
  We propose a differentially private decentralized learning method (termed PrivSGPVR) which employs gradient push with variance reduction and guarantees privacy for each node. Our theoretical analysis shows that, under DP noise with constant variance, PrivGPS-VR achieves a sub-linear convergence rate of $mathcalO (1/sqrtnK)$.
  arXiv  Detail & Related papers  (2024-05-04T11:22:53Z)
• Differential Privacy with Random Projections and Sign Random Projections [37.6593006747285]
  iDP-SignRP is remarkably effective under the setting of individual differential privacy'' (iDP) DP-SignOPORP considerably improves existing algorithms under the standard DP setting.
  arXiv  Detail & Related papers  (2023-05-22T16:33:23Z)
• Near Sample-Optimal Reduction-based Policy Learning for Average Reward MDP [58.13930707612128]
  This work considers the sample complexity of obtaining an $varepsilon$-optimal policy in an average reward Markov Decision Process (AMDP) We prove an upper bound of $widetilde O(H varepsilon-3 ln frac1delta)$ samples per state-action pair, where $H := sp(h*)$ is the span of bias of any optimal policy, $varepsilon$ is the accuracy and $delta$ is the failure probability.
  arXiv  Detail & Related papers  (2022-12-01T15:57:58Z)
• Differentially Private Image Classification from Features [53.75086935617644]
  Leveraging transfer learning has been shown to be an effective strategy for training large models with Differential Privacy. Recent works have found that privately training just the last layer of a pre-trained model provides the best utility with DP.
  arXiv  Detail & Related papers  (2022-11-24T04:04:20Z)
• Normalized/Clipped SGD with Perturbation for Differentially Private Non-Convex Optimization [94.06564567766475]
  DP-SGD and DP-NSGD mitigate the risk of large models memorizing sensitive training data. We show that these two algorithms achieve similar best accuracy while DP-NSGD is comparatively easier to tune than DP-SGD.
  arXiv  Detail & Related papers  (2022-06-27T03:45:02Z)
• Differentially Private Exploration in Reinforcement Learning with Linear Representation [102.17246636801649]
  We first consider the setting of linear-mixture MDPs (Ayoub et al., 2020) (a.k.a. model-based setting) and provide an unified framework for analyzing joint and local differential private (DP) exploration. We further study privacy-preserving exploration in linear MDPs (Jin et al., 2020) (a.k.a. model-free setting) where we provide a $widetildeO(sqrtK/epsilon)$ regret bound for $(epsilon,delta)
  arXiv  Detail & Related papers  (2021-12-02T19:59:50Z)
• Reward-Free Model-Based Reinforcement Learning with Linear Function Approximation [92.99933928528797]
  We study the model-based reward-free reinforcement learning with linear function approximation for episodic Markov decision processes (MDPs) In the planning phase, the agent is given a specific reward function and uses samples collected from the exploration phase to learn a good policy. We show that to obtain an $epsilon$-optimal policy for arbitrary reward function, UCRL-RFE needs to sample at most $tilde O(H4d(H + d)epsilon-2)$ episodes.
  arXiv  Detail & Related papers  (2021-10-12T23:03:58Z)
• Private Stochastic Non-Convex Optimization: Adaptive Algorithms and Tighter Generalization Bounds [72.63031036770425]
  We propose differentially private (DP) algorithms for bound non-dimensional optimization. We demonstrate two popular deep learning methods on the empirical advantages over standard gradient methods.
  arXiv  Detail & Related papers  (2020-06-24T06:01:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.