Corpus ID: 15236510

(Nearly) Optimal Algorithms for Private Online Learning in Full-information and Bandit Settings

@inproceedings{Thakurta2013NearlyOA,
  title={(Nearly) Optimal Algorithms for Private Online Learning in Full-information and Bandit Settings},
  author={Abhradeep Thakurta and Adam D. Smith},
  booktitle={NIPS},
  year={2013}
}
We give differentially private algorithms for a large class of online learning algorithms, in both the full information and bandit settings. Our algorithms aim to minimize a convex loss function which is a sum of smaller convex loss terms, one for each data point. To design our algorithms, we modify the popular mirror descent approach, or rather a variant called follow the approximate leader. The technique leads to the first nonprivate algorithms for private online learning in the bandit… Expand
Optimal Algorithms for Private Online Learning in a Stochastic Environment
TLDR
For the problems of decision-theoretic online learning with stochastic rewards, the first algorithm that achieves an O ( logK ∆min + logK ǫ ) regret bound is presented, where∆min is the minimum mean reward gap. Expand
Private Empirical Risk Minimization, Revisited
TLDR
This paper provides new algorithms and matching lower bounds for private ERM assuming only that each data point’s contribution to the loss function is Lipschitz bounded and that the domain of optimization is bounded, and implies that algorithms from previous work can be used to obtain optimal error rates. Expand
Private Empirical Risk Minimization: Efficient Algorithms and Tight Error Bounds
TLDR
This work provides new algorithms and matching lower bounds for differentially private convex empirical risk minimization assuming only that each data point's contribution to the loss function is Lipschitz and that the domain of optimization is bounded. Expand
Online Learning via the Differential Privacy Lens
TLDR
An algorithmic property called one-step differential stability is developed which facilitates a more refined regret analysis for online learning methods and defines stronger notions of stability that are useful in deriving bounds in partial information settings such as multi-armed bandits and bandits with experts. Expand
(Nearly) Optimal Differentially Private Stochastic Multi-Arm Bandits
TLDR
This work designs algorithms that are i) differentially private, and ii) have regret guarantees that (almost) match the regret guarantees for the best non-private algorithms (e.g., upper confidence bound sampling and Thompson sampling). Expand
Differentially Private Monotone Submodular Maximization Under Matroid and Knapsack Constraints
TLDR
A differentially private (1− κe )-approximation algorithm is proposed, where κ ∈ [0, 1] is the total curvature of the submodular set function, which improves upon prior works in terms of approximation guarantee and query complexity under the same privacy budget. Expand
Private Stochastic Multi-arm Bandits: From Theory to Practice
TLDR
This paper designs algorithms that are i) differentially private, and ii) have regret guarantees that (almost) match the regret guarantees for the best non-private algorithms (e.g., upper confidence bound sampling). Expand
Differentially Private Online Submodular Optimization
TLDR
These are the first algorithms for online submodular minimization that preserve differential privacy under full information feedback and bandit feedback and simulate private online convex optimization using this estimate. Expand
Projection-Free Bandit Optimization with Privacy Guarantees
TLDR
This is the first differentially-private algorithm for projection-free bandit optimization, and in fact its bound matches the best known non-private projection- free algorithm and the bestknown private algorithm, even for the weaker setting when projections are available. Expand
Is Interaction Necessary for Distributed Private Learning?
TLDR
This work asks how much interaction is necessary to optimize convex functions in the local DP model, and provides new algorithms which are either noninteractive or use relatively few rounds of interaction. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 28 REFERENCES
Differentially Private Online Learning
TLDR
This paper provides a general framework to convert the given algorithm into a privacy preserving OCP algorithm with good (sub-linear) regret, and shows that this framework can be used to provide differentially private algorithms for offline learning as well. Expand
Optimal Algorithms for Online Convex Optimization with Multi-Point Bandit Feedback
TLDR
The multi-point bandit setting, in which the player can query each loss function at multiple points, is introduced, and regret bounds that closely resemble bounds for the full information case are proved. Expand
Logarithmic regret algorithms for online convex optimization
TLDR
Several algorithms achieving logarithmic regret are proposed, which besides being more general are also much more efficient to implement, and give rise to an efficient algorithm based on the Newton method for optimization, a new tool in the field. Expand
Private Convex Empirical Risk Minimization and High-dimensional Regression
We consider differentially private algorithms for convex empirical risk minimization (ERM). Differential privacy (Dwork et al., 2006b) is a recently introduced notion of privacy which guarantees thatExpand
Private Convex Optimization for Empirical Risk Minimization with Applications to High-dimensional Regression
TLDR
This work significantly extends the analysis of the “objective perturbation” algorithm of Chaudhuri et al. (2011) for convex ERM problems, and gives the best known algorithms for differentially private linear regression. Expand
Differentially Private Learning with Kernels
TLDR
This paper derives differentially private learning algorithms with provable "utility" or error bounds from the standard learning model of releasing different private predictor using three simple but practical models. Expand
Competing in the Dark: An Efficient Algorithm for Bandit Linear Optimization
TLDR
This work introduces an efficient algorithm for the problem of online linear optimization in the bandit setting which achieves the optimal O∗( √ T ) regret and presents a novel connection between online learning and interior point methods. Expand
Differentially Private Empirical Risk Minimization
TLDR
This work proposes a new method, objective perturbation, for privacy-preserving machine learning algorithm design, and shows that both theoretically and empirically, this method is superior to the previous state-of-the-art, output perturbations, in managing the inherent tradeoff between privacy and learning performance. Expand
Privacy-preserving logistic regression
TLDR
This paper addresses the important tradeoff between privacy and learnability, when designing algorithms for learning from private databases by providing a privacy-preserving regularized logistic regression algorithm based on a new privacy- Preserving technique. Expand
Privacy-preserving statistical estimation with optimal convergence rates
TLDR
It is shown that for a large class of statistical estimators T and input distributions P, there is a differentially private estimator AT with the same asymptotic distribution as T, which implies that AT (X) is essentially as good as the original statistic T(X) for statistical inference, for sufficiently large samples. Expand
...
1
2
3
...