Neural Information Processing Systems http://nips.cc/

Neural Information Processing Systems http://nips.cc/

We sincerely thank all reviewers for their valuable comments. Our responses to the comments are listed below. If we understand it correctly, "the approximate posterior in If so, we do consider the effects of the KL term in the proof. Dir(β), which is a Dirichlet posterior with a Dirichlet prior Dir(α). Does Claim 4.1 rely on a specific distribution (q1 under Weaknesses)?

Furthermore, many important constrained distributions are inherently non-log-concave.

When applied to distributions defined on a constrained space the time-discretization error can dominate when we are near the boundary of the space.

We sincerely thank all reviewers for their valuable comments. Our responses to the comments are listed below. If we understand it correctly, "the approximate posterior in If so, we do consider the effects of the KL term in the proof. Does Claim 4.1 rely on a specific distribution (q1 under As for Gaussian variables (the case in VGAE), we are not sure if the claim still holds. We agree sometimes it might be hard to follow and we shall add the details in the revision.

Besides the Laplace distribution and the Gaussian distribution, there are many more probability distributions which is not well-understood in terms of privacy-preserving property of a random draw -- one of which is the Dirichlet distribution. In this work, we study the inherent privacy of releasing a single draw from a Dirichlet posterior distribution. As a complement to the previous study that provides general theories on the differential privacy of posterior sampling from exponential families, this study focuses specifically on the Dirichlet posterior sampling and its privacy guarantees. With the notion of truncated concentrated differential privacy (tCDP), we are able to derive a simple privacy guarantee of the Dirichlet posterior sampling, which effectively allows us to analyze its utility in various settings. Specifically, we prove accuracy guarantees of private Multinomial-Dirichlet sampling, which is prevalent in Bayesian tasks, and private release of a normalized histogram. In addition, with our results, it is possible to make Bayesian reinforcement learning differentially private by modifying the Dirichlet sampling for state transition probabilities.

Stochastic gradient Markov chain Monte Carlo (SGMCMC) has become a popular method for scalable Bayesian inference. These methods are based on sampling a discrete-time approximation to a continuous time process, such as the Langevin diffusion. When applied to distributions defined on a constrained space the time-discretization error can dominate when we are near the boundary of the space. We demonstrate that because of this, current SGMCMC methods for the simplex struggle with sparse simplex spaces; when many of the components are close to zero. Unfortunately, many popular large-scale Bayesian models, such as network or topic models, require inference on sparse simplex spaces. To avoid the biases caused by this discretization error, we propose the stochastic Cox-Ingersoll-Ross process (SCIR), which removes all discretization error and we prove that samples from the SCIR process are asymptotically unbiased. We discuss how this idea can be extended to target other constrained spaces. Use of the SCIR process within a SGMCMC algorithm is shown to give substantially better performance for a topic model and a Dirichlet process mixture model than existing SGMCMC approaches.

We consider the problem of sampling from constrained distributions, which has posed significant challenges to both non-asymptotic analysis and algorithmic design. We propose a unified framework, which is inspired by the classical mirror descent, to derive novel first-order sampling schemes. We prove that, for a general target distribution with strongly convex potential, our framework implies the existence of a first-order algorithm achieving O~(\epsilon^{-2}d) convergence, suggesting that the state-of-the-art O~(\epsilon^{-6}d^5) can be vastly improved. With the important Latent Dirichlet Allocation (LDA) application in mind, we specialize our algorithm to sample from Dirichlet posteriors, and derive the first non-asymptotic O~(\epsilon^{-2}d^2) rate for first-order sampling. We further extend our framework to the mini-batch setting and prove convergence rates when only stochastic gradients are available. Finally, we report promising experimental results for LDA on real datasets.

Stochastic gradient Markov chain Monte Carlo (SGMCMC) has become a popular method for scalable Bayesian inference. These methods are based on sampling a discrete-time approximation to a continuous time process, such as the Langevin diffusion. When applied to distributions defined on a constrained space the time-discretization error can dominate when we are near the boundary of the space. We demonstrate that because of this, current SGMCMC methods for the simplex struggle with sparse simplex spaces; when many of the components are close to zero. Unfortunately, many popular large-scale Bayesian models, such as network or topic models, require inference on sparse simplex spaces. To avoid the biases caused by this discretization error, we propose the stochastic Cox-Ingersoll-Ross process (SCIR), which removes all discretization error and we prove that samples from the SCIR process are asymptotically unbiased. We discuss how this idea can be extended to target other constrained spaces. Use of the SCIR process within a SGMCMC algorithm is shown to give substantially better performance for a topic model and a Dirichlet process mixture model than existing SGMCMC approaches.