Neural Networks (NNs) have provided state-of-the-art results for many challenging machine learning tasks such as detection, regression and classification across the domains of computer vision, speech recognition and natural language processing. Despite their success, they are often implemented in a frequentist scheme, meaning they are unable to reason about uncertainty in their predictions. This article introduces Bayesian Neural Networks (BNNs) and the seminal research regarding their implementation. Different approximate inference methods are compared, and used to highlight where future research can improve on current methods.

With artificial intelligence systems becoming ubiquitous in our society, its designers will soon have to start to consider its social dimension, as many of these systems will have to interact among them to work efficiently. With this in mind, we propose a decentralized deep reinforcement learning algorithm for the design of cooperative multi-agent systems. The algorithm is based on the hypothesis that highly correlated actions are a feature of cooperative systems, and hence, we propose the insertion of an auxiliary objective of maximization of the mutual information between the actions of agents in the learning problem. Our system is applied to a social dilemma, a problem whose optimal solution requires that agents cooperate to maximize a macroscopic performance function despite the divergent individual objectives of each agent. By comparing the performance of the proposed system to a system without the auxiliary objective, we conclude that the maximization of mutual information among agents promotes the emergence of cooperation in social dilemmas.

The principle of optimism in the face of uncertainty is prevalent throughout sequential decision making problems such as multi-armed bandits and reinforcement learning (RL), often coming with strong theoretical guarantees. However, it remains a challenge to scale these approaches to the deep RL paradigm, which has achieved a great deal of attention in recent years. In this paper, we introduce a tractable approach to optimism via noise augmented Markov Decision Processes (MDPs), which we show can obtain a competitive regret bound: $\tilde{\mathcal{O}}( |\mathcal{S}|H\sqrt{|\mathcal{S}||\mathcal{A}| T } )$ when augmenting using Gaussian noise, where $T$ is the total number of environment steps. This tractability allows us to apply our approach to the deep RL setting, where we rigorously evaluate the key factors for success of optimistic model-based RL algorithms, bridging the gap between theory and practice.

Many democratic countries use district-based elections where there is a "seat" for each district in the governing body. In each district, the party whose candidate gets the maximum number of votes wins the corresponding seat. The result of the election is decided based on the number of seats won by the different parties. The electors (voters) can cast their votes only in the district of their residence. Thus, locations of the electors and boundaries of the districts may severely affect the election result even if the proportion of popular support (number of electors) of different parties remains unchanged. This has led to significant amount of research on whether the districts may be redrawn or electors may be moved to maximize seats for a particular party. In this paper, we frame the spatial distribution of electors in a probabilistic setting, and explore different models to capture the intra-district polarization of electors in favour of a party, or the spatial concentration of supporters of different parties. Our models are inspired by elections in India, where supporters of different parties tend to be concentrated in certain districts. We show with extensive simulations that our model can capture different statistical properties of real elections held in India. We frame parameter estimation problems to fit our models to the observed election results. Since analytical calculation of the likelihood functions are infeasible for our complex models, we use Likelihood-free Inference methods under the Approximate Bayesian Computation framework. Since this approach is highly time-consuming, we explore how supervised regression using Logistic Regression or Deep Neural Networks can be used to speed it up. We also explore how the election results can change by varying the spatial distributions of the voters, even when the proportions of popular support of the parties remain constant.

Bayesian meta-learning enables robust and fast adaptation to new tasks with uncertainty assessment. The key idea behind Bayesian meta-learning is empirical Bayes inference of hierarchical model. In this work, we extend this framework to include a variety of existing methods, before proposing our variant based on gradient-EM algorithm. Our method improves computational efficiency by avoiding back-propagation computation in the meta-update step, which is exhausting for deep neural networks. Furthermore, it provides flexibility to the inner-update optimization procedure by decoupling it from meta-update. Experiments on sinusoidal regression, few-shot image classification, and policy-based reinforcement learning show that our method not only achieves better accuracy with less computation cost, but is also more robust to uncertainty.

The model uncertainty obtained by variational Bayesian inference with Monte Carlo dropout is prone to miscalibration. In this paper, different logit scaling methods are extended to dropout variational inference to recalibrate model uncertainty. Expected uncertainty calibration error (UCE) is presented as a metric to measure miscalibration. The effectiveness of recalibration is evaluated on CIFAR-10/100 and SVHN for recent CNN architectures. Experimental results show that logit scaling considerably reduce miscalibration by means of UCE. Well-calibrated uncertainty enables reliable rejection of uncertain predictions and robust detection of out-of-distribution data.

We consider an inverse reinforcement learning problem involving us versus an enemy radar equipped with a Bayesian tracker. By observing the emissions of the enemy radar,how can we identify if the radar is cognitive (constrained utility maximizer)? Given the observed sequence of actions taken by the enemy's radar, we consider three problems: (i) Are the enemy radar's actions (waveform choice, beam scheduling) consistent with constrained utility maximization? If so how can we estimate the cognitive radar's utility function that is consistent with its actions. We formulate and solve the problem in terms of the spectra (eigenvalues) of the state and observation noise covariance matrices, and the algebraic Riccati equation. (ii) How to construct a statistical test for detecting a cognitive radar (constrained utility maximization) when we observe the radar's actions in noise or the radar observes our probe signal in noise? We propose a statistical detector with a tight Type-II error bound. (iii) How can we optimally probe (interrogate) the enemy's radar by choosing our state to minimize the Type-II error of detecting if the radar is deploying an economic rational strategy, subject to a constraint on the Type-I detection error? We present a stochastic optimization algorithm to optimize our probe signal. The main analysis framework used in this paper is that of revealed preferences from microeconomics.

An interesting development in automatic visual recognition has been the emergence of tasks where it is not possible to assign ground truth labels to images, yet still feasible to collect annotations that reflect human judgements about them. Such tasks include subjective visual attribute assignment and the labeling of ambiguous scenes. Machine learning-based predictors for these tasks rely on supervised training that models the behavior of the annotators, e.g., what would the average person's judgement be for an image? A key open question for this type of work, especially for applications where inconsistency with human behavior can lead to ethical lapses, is how to evaluate the uncertainty of trained predictors. Given that the real answer is unknowable, we are left with often noisy judgements from human annotators to work with. In order to account for the uncertainty that is present, we propose a relative Bayesian framework for evaluating predictors trained on such data. The framework specifies how to estimate a predictor's uncertainty due to the human labels by approximating a conditional distribution and producing a credible interval for the predictions and their measures of performance. The framework is successfully applied to four image classification tasks that use subjective human judgements: facial beauty assessment using the SCUT-FBP5500 dataset, social attribute assignment using data from TestMyBrain.org,

Inverse reinforcement learning (IRL) aims to estimate the reward function of optimizing agents by observing their response (estimates or actions). This paper considers IRL when noisy estimates of the gradient of a reward function generated by multiple stochastic gradient agents are observed. We present a generalized Langevin dynamics algorithm to estimate the reward function $R(\theta)$; specifically, the resulting Langevin algorithm asymptotically generates samples from the distribution proportional to $\exp(R(\theta))$. The proposed IRL algorithms use kernel-based passive learning schemes. We also construct multi-kernel passive Langevin algorithms for IRL which are suitable for high dimensional data. The performance of the proposed IRL algorithms are illustrated on examples in adaptive Bayesian learning, logistic regression (high dimensional problem) and constrained Markov decision processes. We prove weak convergence of the proposed IRL algorithms using martingale averaging methods. We also analyze the tracking performance of the IRL algorithms in non-stationary environments where the utility function $R(\theta)$ jump changes over time as a slow Markov chain.

Reliably transferring treatment policies learned in one clinical environment to another is currently limited by challenges related to domain shift. In this paper we address off-policy learning for sequential decision making under domain shift -- a scenario susceptible to catastrophic overconfidence -- which is highly relevant to a high-stakes clinical settings where the target domain may also be data-scarce. We propose a two-fold counterfactual regularization procedure to improve off-policy learning, addressing domain shift and data scarcity. First, we utilize an informative prior derived from a data-rich source environment to indirectly improve drawing counterfactual example observations. Then, these samples are then used to learn a policy for the target domain, regularized by the source policy through KL-divergence. In simulated sepsis treatment, our counterfactual policy transfer procedure significantly improves the performance of a learned treatment policy.