These notes review six lectures given by Prof. Andrea Montanari on the topic of statistical estimation for linear models. The first two lectures cover the principles of signal recovery from linear measurements in terms of minimax risk. Subsequent lectures demonstrate the application of these principles to several practical problems in science and engineering. Specifically, these topics include denoising of error-laden signals, recovery of compressively sensed signals, reconstruction of low-rank matrices, and also the discovery of hidden cliques within large networks. These are notes from the lecture of Andrea Montanari given at the autumn school "Statistical Physics, Optimization, Inference, and Message-Passing Algorithms", that took place in Les Houches, France from Monday September 30th, 2013, till Friday October 11th, 2013.

We derive relations between theoretical properties of restricted Boltzmann machines (RBMs), popular machine learning models which form the building blocks of deep learning models, and several natural notions from discrete mathematics and convex geometry. We give implications and equivalences relating RBM-representable probability distributions, perfectly reconstructible inputs, Hamming modes, zonotopes and zonosets, point configurations in hyperplane arrangements, linear threshold codes, and multi-covering numbers of hypercubes. As a motivating application, we prove results on the relative representational power of mixtures of product distributions and products of mixtures of pairs of product distributions (RBMs) that formally justify widely held intuitions about distributed representations. In particular, we show that a mixture of products requiring an exponentially larger number of parameters is needed to represent the probability distributions which can be obtained as products of mixtures.

Gibbs sampling is a workhorse for Bayesian inference but has several limitations when used for parameter estimation, and is often much slower than non-sampling inference methods. SAME (State Augmentation for Marginal Estimation) \cite{Doucet99,Doucet02} is an approach to MAP parameter estimation which gives improved parameter estimates over direct Gibbs sampling. SAME can be viewed as cooling the posterior parameter distribution and allows annealed search for the MAP parameters, often yielding very high quality (lower loss) estimates. But it does so at the expense of additional samples per iteration and generally slower performance. On the other hand, SAME dramatically increases the parallelism in the sampling schedule, and is an excellent match for modern (SIMD) hardware. In this paper we explore the application of SAME to graphical model inference on modern hardware. We show that combining SAME with factored sample representation (or approximation) gives throughput competitive with the fastest symbolic methods, but with potentially better quality. We describe experiments on Latent Dirichlet Allocation, achieving speeds similar to the fastest reported methods (online Variational Bayes) and lower cross-validated loss than other LDA implementations. The method is simple to implement and should be applicable to many other models.

Ashley Petersen, Daniela Witten†, and Noah Simon ‡ Department of Biostatistics, University of Washington, Seattle W A 98195 March 29, 2018 We consider the problem of predicting an outcome variable usingp covariates that are measured onn independent observations, in the setting in which flexible and interpretable fits are desirable. We propose the fused lasso additive model (FLAM), in which each additive function is estimated to be piecewise constant with a small number of adaptively-chosen knots. FLAM is the solution to a convex optimization problem, for which a simple algorithm with guaranteed convergence to the global optimum is provided. FLAM is shown to be consistent in high dimensions, and an unbiased estimator of its degrees of freedom is proposed. We evaluate the performance of FLAM in a simulation study and on two data sets. Keywords: additive model, feature selection, high-dimensional, nonparametric regression, piecewise constant, sparsity 1 Introduction In this paper, we consider the task of predicting a response variable usingp features measured on n independent observations. In this paper, we propose a method that balances the tradeoff between inter-pretability and flexibility, while also allowing for sparsity in high dimensions whenp n . It selects a subset of features to include in the model, and for these features it fits piecewise constant functions with knots that are chosen adaptively based on the data. We now introduce some notation. We letX denote ann p matrix, for whichx j is the j th column (feature), and for which thei th element (observation) isx ij .

Particle Metropolis-Hastings (PMH) allows for Bayesian parameter inference in nonlinear state space models by combining Markov chain Monte Carlo (MCMC) and particle filtering. The latter is used to estimate the intractable likelihood. In its original formulation, PMH makes use of a marginal MCMC proposal for the parameters, typically a Gaussian random walk. However, this can lead to a poor exploration of the parameter space and an inefficient use of the generated particles. We propose a number of alternative versions of PMH that incorporate gradient and Hessian information about the posterior into the proposal. This information is more or less obtained as a byproduct of the likelihood estimation. Indeed, we show how to estimate the required information using a fixed-lag particle smoother, with a computational cost growing linearly in the number of particles. We conclude that the proposed methods can: (i) decrease the length of the burn-in phase, (ii) increase the mixing of the Markov chain at the stationary phase, and (iii) make the proposal distribution scale invariant which simplifies tuning.

Simplifying expressions is important to make numerical integration of large expressions from High Energy Physics tractable. To this end, Horner's method can be used. Finding suitable Horner schemes is assumed to be hard, due to the lack of local heuristics. Recently, MCTS was reported to be able to find near optimal schemes. However, several parameters had to be fine-tuned manually. In this work, we investigate the state space properties of Horner schemes and find that the domain is relatively flat and contains only a few local minima. As a result, the Horner space is appropriate to be explored by Stochastic Local Search (SLS), which has only two parameters: the number of iterations (computation time) and the neighborhood structure. We found a suitable neighborhood structure, leaving only the allowed computation time as a parameter. We performed a range of experiments. The results obtained by SLS are similar or better than those obtained by MCTS. Furthermore, we show that SLS obtains the good results at least 10 times faster. Using SLS, we can speed up numerical integration of many real-world large expressions by at least a factor of 24. For High Energy Physics this means that numerical integrations that took weeks can now be done in hours.

When integrating information coming from different sources, a distinction is made between revision [13, 5, 14, 28, 6] (new information more reliable than old) and merging [22, 4, 18] (same reliability). More generally, priorities or weights are assigned to the sources to indicate their reliability [26, 27, 30, 7]. Measures and aggregation functions allow for fine-grained policies of integration [16, 11, 18]. Families of operators are then defined, all depending in a way or another from the relative reliability of the sources. The two basic cases of non-iterated revision and merging result from giving priority to the new information or the same to all pieces of information to be incorporated, respectively. The strenght of information sources has been studied in the field of cognitive psychology, where it was determined to depend on the order in which the information is given [32], on the size of the group generating it [25] and other social factors [31]. The first time merging is done, the relative reliability of the pieces of information to be integrated cannot come other than from sources external to the merging process. However, subsequent mergings may then take advantage from the previous results.

In this study, we enrich the framework of nonparametric kernel Bayesian inference via the flexible incorporation of certain probabilistic models, such as additive Gaussian noise models. Nonparametric inference expressed in terms of kernel means, which is called kernel Bayesian inference, has been studied using basic rules such as the kernel sum rule (KSR), kernel chain rule, kernel product rule, and kernel Bayes' rule (KBR). However, the current framework used for kernel Bayesian inference deals only with nonparametric inference and it cannot allow inference when combined with probabilistic models. In this study, we introduce a novel KSR, called model-based KSR (Mb-KSR), which exploits the knowledge obtained from some probabilistic models of conditional distributions. The incorporation of Mb-KSR into nonparametric kernel Bayesian inference facilitates more flexible kernel Bayesian inference than nonparametric inference. We focus on combinations of Mb-KSR, Non-KSR, and KBR, and we propose a filtering algorithm for state space models, which combines nonparametric learning of the observation process using kernel means and additive Gaussian noise models of the transition dynamics. The idea of the Mb-KSR for additive Gaussian noise models can be extended to more general noise model cases, including a conjugate pair with a positive-definite kernel and a probabilistic model.

A survey of existing methods for stopping active learning (AL) reveals the needs for methods that are: more widely applicable; more aggressive in saving annotations; and more stable across changing datasets. A new method for stopping AL based on stabilizing predictions is presented that addresses these needs. Furthermore, stopping methods are required to handle a broad range of different annotation/performance tradeoff valuations. Despite this, the existing body of work is dominated by conservative methods with little (if any) attention paid to providing users with control over the behavior of stopping methods. The proposed method is shown to fill a gap in the level of aggressiveness available for stopping AL and supports providing users with control over stopping behavior.

Although recovering an Euclidean distance matrix from noisy observations is a common problem in practice, how well this could be done remains largely unknown. To fill in this void, we study a simple distance matrix estimate based upon the so-called regularized kernel estimate. We show that such an estimate can be characterized as simply applying a constant amount of shrinkage to all observed pairwise distances. This fact allows us to establish risk bounds for the estimate implying that the true distances can be estimated consistently in an average sense as the number of objects increases. In addition, such a characterization suggests an efficient algorithm to compute the distance matrix estimator, as an alternative to the usual second order cone programming known not to scale well for large problems. Numerical experiments and an application in visualizing the diversity of Vpu protein sequences from a recent HIV-1 study further demonstrate the practical merits of the proposed method.