Likelihood-free inference for simulator-based statistical models has recently grown rapidly from its infancy to a useful tool for practitioners. However, models with more than a very small number of parameters as the target of inference have remained an enigma, in particular for the approximate Bayesian computation (ABC) community. To advance the possibilities for performing likelihood-free inference in high-dimensional parameter spaces, here we introduce an extension of the popular Bayesian optimisation based approach to approximate discrepancy functions in a probabilistic manner which lends itself to an efficient exploration of the parameter space. Our method achieves computational scalability by using separate acquisition procedures for the discrepancies defined for different parameters. These efficient high-dimensional simulation acquisitions are combined with exponentiated loss-likelihoods to provide a misspecification-robust characterisation of the marginal posterior distribution for all model parameters. The method successfully performs computationally efficient inference in a 100-dimensional space on canonical examples and compares favourably to existing Copula-ABC methods. We further illustrate the potential of this approach by fitting a bacterial transmission dynamics model to daycare centre data, which provides biologically coherent results on the strain competition in a 30-dimensional parameter space.

We consider the problem of estimating a ranking on a set of items from noisy pairwise comparisons given item features. We address the fact that pairwise comparison data often reflects irrational choice, e.g. intransitivity. Our key observation is that two items compared in isolation from other items may be compared based on only a salient subset of features. Formalizing this framework, we propose the "salient feature preference model" and prove a sample complexity result for learning the parameters of our model and the underlying ranking with maximum likelihood estimation. We also provide empirical results that support our theoretical bounds and illustrate how our model explains systematic intransitivity. Finally we demonstrate strong performance of maximum likelihood estimation of our model on both synthetic data and two real data sets: the UT Zappos50K data set and comparison data about the compactness of legislative districts in the US.

This paper studies the statistical theory of batch data reinforcement learning with function approximation. Consider the off-policy evaluation problem, which is to estimate the cumulative value of a new target policy from logged history generated by unknown behavioral policies. We study a regression-based fitted Q iteration method, and show that it is equivalent to a model-based method that estimates a conditional mean embedding of the transition operator. We prove that this method is information-theoretically optimal and has nearly minimal estimation error. In particular, by leveraging contraction property of Markov processes and martingale concentration, we establish a finite-sample instance-dependent error upper bound and a nearly-matching minimax lower bound. The policy evaluation error depends sharply on a restricted $\chi^2$-divergence over the function class between the long-term distribution of the target policy and the distribution of past data. This restricted $\chi^2$-divergence is both instance-dependent and function-class-dependent. It characterizes the statistical limit of off-policy evaluation. Further, we provide an easily computable confidence bound for the policy evaluator, which may be useful for optimistic planning and safe policy improvement.

In some applications, acquiring covariates comes at a cost which is not negligible. For example in the medical domain, in order to classify whether a patient has diabetes or not, measuring glucose tolerance can be expensive. Assuming that the cost of each covariate, and the cost of misclassification can be specified by the user, our goal is to minimize the (expected) total cost of classification, i.e. the cost of misclassification plus the cost of the acquired covariates. We formalize this optimization goal using the (conditional) Bayes risk and describe the optimal solution using a recursive procedure. Since the procedure is computationally infeasible, we consequently introduce two assumptions: (1) the optimal classifier can be represented by a generalized additive model, (2) the optimal sets of covariates are limited to a sequence of sets of increasing size. We show that under these two assumptions, a computationally efficient solution exists. Furthermore, on several medical datasets, we show that the proposed method achieves in most situations the lowest total costs when compared to various previous methods. Finally, we weaken the requirement on the user to specify all misclassification costs by allowing the user to specify the minimally acceptable recall (target recall). Our experiments confirm that the proposed method achieves the target recall while minimizing the false discovery rate and the covariate acquisition costs better than previous methods.

Matrix factorization (MF) has been widely applied to collaborative filtering in recommendation systems. Its Bayesian variants can derive posterior distributions of user and item embeddings, and are more robust to sparse ratings. However, the Bayesian methods are restricted by their update rules for the posterior parameters due to the conjugacy of the priors and the likelihood. Neural networks can potentially address this issue by capturing complex mappings between the posterior parameters and the data. In this paper, we propose a variational auto-encoder based Bayesian MF framework. It leverages not only the data but also the information from the embeddings to approximate their joint posterior distribution. The approximation is an iterative procedure with cross feedback of user and item embeddings to the others' encoders. More specifically, user embeddings sampled in the previous iteration, alongside their ratings, are fed back into the item-side encoders to compute the posterior parameters for the item embeddings in the current iteration, and vice versa. The decoder network then reconstructs the data using the MF with the currently re-sampled user and item embeddings. We show the effectiveness of our framework in terms of reconstruction errors across five real-world datasets. We also perform ablation studies to illustrate the importance of the cross feedback component of our framework in lowering the reconstruction errors and accelerating the convergence.

We present new results on the classical algorithm of variable elimination, which underlies many algorithms including for probabilistic inference. The results relate to exploiting functional dependencies, allowing one to perform inference and learning efficiently on models that have very large treewidth. The highlight of the advance is that it works with standard (dense) factors, without the need for sparse factors or techniques based on knowledge compilation that are commonly utilized. This is significant as it permits a direct implementation of the improved variable elimination algorithm using tensors and their operations, leading to extremely efficient implementations especially when learning model parameters. Moreover, the proposed technique does not require knowledge of the specific functional dependencies, only that they exist, so can be used when learning these dependencies. We illustrate the efficacy of our proposed algorithm by compiling Bayesian network queries into tensor graphs and then learning their parameters from labeled data using a standard tool for tensor computation.

We introduce Deep Sigma Point Processes, a class of parametric models inspired by the compositional structure of Deep Gaussian Processes (DGPs). Deep Sigma Point Processes (DSPPs) retain many of the attractive features of (variational) DGPs, including mini-batch training and predictive uncertainty that is controlled by kernel basis functions. Importantly, since DSPPs admit a simple maximum likelihood inference procedure, the resulting predictive distributions are not degraded by any posterior approximations. In an extensive empirical comparison on univariate and multivariate regression tasks we find that the resulting predictive distributions are significantly better calibrated than those obtained with other probabilistic methods for scalable regression, including variational DGPs--often by as much as a nat per datapoint.

Bayesian reward learning from demonstrations enables rigorous safety and uncertainty analysis when performing imitation learning. However, Bayesian reward learning methods are typically computationally intractable for complex control problems. We propose a highly efficient Bayesian reward learning algorithm that scales to high-dimensional imitation learning problems by first pre-training a low-dimensional feature encoding via self-supervised tasks and then leveraging preferences over demonstrations to perform fast Bayesian inference. We evaluate our proposed approach on the task of learning to play Atari games from demonstrations, without access to the game score. For Atari games our approach enables us to generate 100,000 samples from the posterior over reward functions in only 5 minutes using a personal laptop. Furthermore, our proposed approach achieves comparable or better imitation learning performance than state-of-the-art methods that only find a point estimate of the reward function. Finally, we show that our approach enables efficient high-confidence policy performance bounds. We show that these high-confidence performance bounds can be used to rank the performance and risk of a variety of evaluation policies, despite not having samples of the reward function. We also show evidence that high-confidence performance bounds can be used to detect reward hacking in complex imitation learning problems.

Recent advances in machine learning highlight successes on a small set of tasks where a large number of labeled examples have been collected and exploited. These include image classification with 1.2 million labeled examples Deng et al. (2009) and French-English machine translation with 40 million paired sentences Bojar et al. (2014). For common tasks, however, collecting clean labels is costly, as they require human expertise (as in medical imaging) or physical interactions (as in robotics), for example. Thus collected real-world datasets follow a long-tailed distribution, in which a dominant set of tasks only have a small number of training examples Wang et al. (2017). Inspired by human ingenuity in quickly solving novel problems by leveraging prior experience, meta-learning approaches aim to jointly learn from past experience to quickly adapt to new tasks with little available data Schmidhuber (1987); Thrun & Pratt (2012). This has had a significant impact in few-shot supervised learning, where each task is associated with only a few training examples. By leveraging structural similarities among those tasks, one can achieve accuracy far greater than what can be achieved for each task in isolation Finn et al. (2017); Ravi & Larochelle (2016); Koch et al. (2015); Oreshkin et al. (2018); Triantafillou et al. (2019); Rusu et al. (2018). The success of such approaches hinges on the following fundamental question: When can we jointly train small data tasks to achieve the accuracy of large data tasks? We investigate this tradeoff under a canonical scenario where the tasks are linear regressions in d-dimensions and the regression parameters are drawn i.i.d.

Statistical inference using pairwise comparison data has been an effective approach to analyzing complex and sparse networks. In this paper we propose a general framework for modeling the mutual interaction in a probabilistic network, which enjoys ample flexibility in terms of parametrization. Within this set-up, we establish that the maximum likelihood estimator (MLE) for the latent scores of the subjects is uniformly consistent under a near-minimal condition on network sparsity. This condition is sharp in terms of the leading order asymptotics describing the sparsity. The proof utilizes a novel chaining technique based on the error-induced metric as well as careful counting of comparison graph structures. Our results guarantee that the MLE is a valid estimator for inference in large-scale comparison networks where data is asymptotically deficient. Numerical simulations are provided to complement the theoretical analysis.