In this paper the authors presented a sequential variational inference algorithm for Dirichlet process mixture models. The authors used a posterior characterization of normalized random measures with independent increments as the basis for a variational distribution that was then used on a sequential decomposition of the posterior. The algorithm was demonstrated on a Gaussian mixture model applied to real and synthetic data, and a non-conjugate DP mixture of Dirichlet distributions to cluster text documents. This is a nice paper that aims to make a particular Bayesian nonparametric model useful to analyze massive data sets. The overall presentation is clear and well-motivated.

Summary: This paper addresses the problem of conditional density estimation with a high dimensional input space (p n), an important problems as most (if not all) current models for nonparametric conditional density estimation do not scale to high-dimensions. Moreover, datasets with high dimensional inputs but relatively small sample sizes are becoming increasingly common. The model for the conditional density f(y x) is defined in three stages. First, a tree structure is defined over the input space. Second, given the tree structure, C_{j,k}, the k th partition of the X space at scale j, is mapped to a lower dimensional space.

Nonparametric estimation of the conditional distribution of a response given highdimensional features is a challenging problem. It is important to allow not only the mean but also the variance and shape of the response density to change flexibly with features, which are massive-dimensional. We propose a multiscale dictionary learning model, which expresses the conditional response density as a convex combination of dictionary densities, with the densities used and their weights dependent on the path through a tree decomposition of the feature space. A fast graph partitioning algorithm is applied to obtain the tree decomposition, with Bayesian methods then used to adaptively prune and average over different sub-trees in a soft probabilistic manner.

Measuring similarity is crucial to many learning tasks. To this end, metric learning has been the dominant paradigm. However, similarity is a richer and broader notion than what metrics entail. For example, similarity can arise from the process of aggregating the decisions of multiple latent components, where each latent component compares data in its own way by focusing on a different subset of features. In this paper, we propose Similarity Component Analysis (SCA), a probabilistic graphical model that discovers those latent components from data.

The paper proposes a Bayesian inference in Monte-Carlo tree search (MCTS) with Thompson sampling based action-selection strategy, called Dirichlet-NormalGamma MCTS (DNG-MTCS) algorithm. The method approximates the accumulated reward of following the current policy from a state, X_{s,\pi(s)}, by the normal distribution with the NromalGamma distribution prior. The state transition probabilities are estimated via Dirichlet distributions. Action-selection strategy is based on Thompson sampling approach, where the expected cumulative reward for each action is computed with the parametric distribution with parameters drawn from the posterior distributions and then the action with the highest expectation is selected. The authors apply the proposed method to several benchmark tasks and showed that the method can converge (slightly) faster than the UCT algorithm. Theoretical properties about convergence are also provided.

Monte-Carlo tree search (MCTS) has been drawing great interest in recent years for planning and learning under uncertainty. One of the key challenges is the trade-off between exploration and exploitation. To address this, we present a novel approach for MCTS using Bayesian mixture modeling and inference based Thompson sampling and apply it to the problem of online planning in MDPs. Our algorithm, named Dirichlet-NormalGamma MCTS (DNG-MCTS), models the uncertainty of the accumulated reward for actions in the search tree as a mixture of Normal distributions. We perform inferences on the mixture in Bayesian settings by choosing conjugate priors in the form of combinations of Dirichlet and NormalGamma distributions and select the best action at each decision node using Thompson sampling. Experimental results confirm that our algorithm advances the state-of-the-art UCT approach with better values on several benchmark problems.

Markov Decision Processes (MDPs) are extremely useful for modeling and solving sequential decision making problems. Graph-based MDPs provide a compact representation for MDPs with large numbers of random variables. However, the complexity of exactly solving a graph-based MDP usually grows exponentially in the number of variables, which limits their application. We present a new variational framework to describe and solve the planning problem of MDPs, and derive both exact and approximate planning algorithms. In particular, by exploiting the graph structure of graph-based MDPs, we propose a factored variational value iteration algorithm in which the value function is first approximated by the multiplication of local-scope value functions, then solved by minimizing a Kullback-Leibler (KL) divergence. The KL divergence is optimized using the belief propagation algorithm, with complexity exponential in only the cluster size of the graph. Experimental comparison on different models shows that our algorithm outperforms existing approximation algorithms at finding good policies.

In this work we develop efficient methods for learning random MAP predictors for structured label problems. In particular, we construct posterior distributions over perturbations that can be adjusted via stochastic gradient methods. We show that any smooth posterior distribution would suffice to define a smooth PAC-Bayesian risk bound suitable for gradient methods. In addition, we relate the posterior distributions to computational properties of the MAP predictors. We suggest multiplicative posteriors to learn super-modular potential functions that accompany specialized MAP predictors such as graph-cuts. We also describe label-augmented posterior models that can use efficient MAP approximations, such as those arising from linear program relaxations.

We describe a class of algorithms for amortized inference in Bayesian networks. In this setting, we invest computation upfront to support rapid online inference for a wide range of queries. Our approach is based on learning an inverse factorization of a model's joint distribution: a factorization that turns observations into root nodes. Our algorithms accumulate information to estimate the local conditional distributions that constitute such a factorization. These stochastic inverses can be used to invert each of the computation steps leading to an observation, sampling backwards in order to quickly find a likely explanation. We show that estimated inverses converge asymptotically in number of (prior or posterior) training samples. To make use of inverses before convergence, we describe the Inverse MCMC algorithm, which uses stochastic inverses to make block proposals for a Metropolis-Hastings sampler. We explore the efficiency of this sampler for a variety of parameter regimes and Bayes nets.

This paper addresses the scalability of symbolic planning under uncertainty with factored states and actions. Our first contribution is a symbolic implementation of Modified Policy Iteration (MPI) for factored actions that views policy evaluation as policy-constrained value iteration (VI). Unfortunately, a naïve approach to enforce policy constraints can lead to large memory requirements, sometimes making symbolic MPI worse than VI. We address this through our second and main contribution, symbolic Opportunistic Policy Iteration (OPI), which is a novel convergent algorithm lying between VI and MPI, that applies policy constraints if it does not increase the size of the value function representation, and otherwise performs VI backups. We also give a memory bounded version of this algorithm allowing a space-time tradeoff. Empirical results show significantly improved scalability over state-of-the-art symbolic planners.