Methods and materials are described for employing a human-shaped robot as a lecturer in automated remote instruction. Video segments from the stimuli of a 2 (participant substrate: VR or non-VR) x 2 (robot embodiment: copresent or screen) balanced between-participants experiment are provided. In each condition, a robot delivers the content for a lecture on the nutrition of carbohydrates. The robot uses identical speech and body movement while the same set of slides plays on an adjacent computer, thereby controlling for such factors as educational content, robot appearance and robot size. The experiment employs Aldebaran’s 25-degrees-of-freedom Nao as the robot and the Oculus Rift as the immersive VR system. The lecture speech and slides were obtained with permission from a Mandarin Chinese-language online course and translated into English. The setup for different delivery modes for automated remote instruction are illustrated using a robot delivering foreign language online content. These methods support the design and evaluation of robots that perform the role of lecturer.

The increasing mere presence of robots in everyday life does not automatically result in gradual acceptance of these systems by human users. Over the past years, we have conducted several studies with the goal to provide insight into the long-term process of social robots in domestic environments. This paper presents our overall conclusions from the combined findings of our multiple studies on social robot acceptance. We will provide insights from a user’s perspective of what makes robots social, describe a phased framework of the long-term process of robot acceptance, present some key factors for social robot acceptance, offer guidelines to build better sociable robots, and provide some recommendations for conducting research in domestic environments. With sharing our experiences with conducting (long-term) user studies in domestic environments, we aim to serve to push this sub-field of HRI in real-world contexts forward and thereby the community at large.

We present a new algorithm called PIEM to approximately solve for the policy of an infinite-horizon decentralized partially observable Markov decision process (DEC-POMDP). The algorithm uses expectation maximization (EM) only in the step of policy improvement, with policy evaluation achieved by solving the Bellman's equation in terms of finite state controllers (FSCs). This marks a key distinction of PIEM from the previous EM algorithm of (Kumar and Zilberstein, 2010), i.e., PIEM directly operates on a DEC-POMDP without transforming it into a mixture of dynamic Bayes nets. Thus, PIEM precisely maximizes the value function, avoiding complicated forward/backward message passing and the corresponding computational and memory cost. To overcome local optima, we follow (Pajarinen and Peltonen, 2011) to solve the DEC-POMDP for a finite length horizon and use the resulting policy graph to initialize the FSCs. We solve the finite-horizon problem using a modified point-based policy generation (PBPG) algorithm, in which a closed-form solution is provided which was previously found by linear programming in the original PBPG. Experimental results on benchmark problems show that the proposed algorithms compare favorably to state-of-the-art methods.

Control of the air in envisioned large-scale battles against near-peer adversaries will require revolutionary new approaches to airborne mission management, where decision authority and platform autonomy are dynamically delegated and functional roles and combat capabilities are assigned across multiple distributed tiers of platforms and human operators. System capabilities range from traditional airborne battle managers, to manned tactical aviators, to autonomous unmanned aerial systems. Due to the overwhelming complexity, human operators will require the assistance of advanced autonomy decision aids with new mechanisms for operator supervision and management of teams of manned and unmanned systems. In this paper we describe a conceptual distributed mission management approach that employs novel human-automation teaming constructs to address the complexity of envisioned operations in highly contested environments. We then discuss a cognitive engineering approach to designing roleand task-tailored human machine interfaces between humans and the autonomous systems. We conclude with a discussion of multi-level evaluation approaches for experimentation.

In a time of pervasive and increasingly transparent computing, humans will interact with information objects and less and less with the computing devices that define them. Artificial Intelligence (AI) will be the proxy for humans’ interaction with information. Because interaction creates opportunities for error, the trend towards AI-augmented human information interaction (HII) will mandate an increased emphasis on cognition-oriented information science research and new ways of thinking about errors and error handling. A review of HII and its relationship to AI is presented, with a focus on errors in this context.

We present a hybrid continuum-atomistic scheme which combines molecular dynamics (MD) simulations with on-the-fly machine learning techniques for the accurate and efficient prediction of multiscale fluidic systems. By using a Gaussian process as a surrogate model for the computationally expensive MD simulations, we use Bayesian inference to predict the system behaviour at the atomistic scale, purely by consideration of the macroscopic inputs and outputs. Whenever the uncertainty of this prediction is greater than a predetermined acceptable threshold, a new MD simulation is performed to continually augment the database, which is never required to be complete. This provides a substantial enhancement to the current generation of hybrid methods, which often require many similar atomistic simulations to be performed, discarding information after it is used once. We apply our hybrid scheme to nano-confined unsteady flow through a high-aspect-ratio converging-diverging channel, and make comparisons between the new scheme and full MD simulations for a range of uncertainty thresholds and initial databases. For low thresholds, our hybrid solution is highly accurate\,---\,within the thermal noise of a full MD simulation. As the uncertainty threshold is raised, the accuracy of our scheme decreases and the computational speed-up increases (relative to a full MD simulation), enabling the compromise between precision and efficiency to be tuned. The speed-up of our hybrid solution ranges from an order of magnitude, with no initial database, to cases where an extensive initial database ensures no new MD simulations are required.

Stochastic Gradient Descent (SGD) is one of the most widely used techniques for online optimization in machine learning. In this work, we accelerate SGD by adaptively learning how to sample the most useful training examples at each time step. First, we show that SGD can be used to learn the best possible sampling distribution of an importance sampling estimator. Second, we show that the sampling distribution of an SGD algorithm can be estimated online by incrementally minimizing the variance of the gradient. The resulting algorithm -- called Adaptive Weighted SGD (AW-SGD) -- maintains a set of parameters to optimize, as well as a set of parameters to sample learning examples. We show that AW-SGD yields faster convergence in three different applications: (i) image classification with deep features, where the sampling of images depends on their labels, (ii) matrix factorization, where rows and columns are not sampled uniformly, and (iii) reinforcement learning, where the optimized and exploration policies are estimated at the same time, where our approach corresponds to an off-policy gradient algorithm.

Capturing the dependence structure of multivariate extreme events is a major concern in many fields involving the management of risks stemming from multiple sources, e.g. portfolio monitoring, insurance, environmental risk management and anomaly detection. One convenient (non-parametric) characterization of extremal dependence in the framework of multivariate Extreme Value Theory (EVT) is the angular measure, which provides direct information about the probable 'directions' of extremes, that is, the relative contribution of each feature/coordinate of the 'largest' observations. Modeling the angular measure in high dimensional problems is a major challenge for the multivariate analysis of rare events. The present paper proposes a novel methodology aiming at exhibiting a sparsity pattern within the dependence structure of extremes. This is done by estimating the amount of mass spread by the angular measure on representative sets of directions, corresponding to specific sub-cones of $R^d\_+$. This dimension reduction technique paves the way towards scaling up existing multivariate EVT methods. Beyond a non-asymptotic study providing a theoretical validity framework for our method, we propose as a direct application a --first-- anomaly detection algorithm based on multivariate EVT. This algorithm builds a sparse 'normal profile' of extreme behaviours, to be confronted with new (possibly abnormal) extreme observations. Illustrative experimental results provide strong empirical evidence of the relevance of our approach.

ABSTRACT Many of the current state-of-the-art Large V ocabulary Continuous Speech Recognition Systems (L VCSR) are hybrids of neural networks and Hidden Markov Models (HMMs). Most of these systems contain separate components that deal with the acoustic modelling, language modelling and sequence decoding. We investigate a more direct approach in which the HMM is replaced with a Recurrent Neural Network (RNN) that performs sequence prediction directly at the character level. Alignment between the input features and the desired character sequence is learned automatically by an attention mechanism built into the RNN. For each predicted character, the attention mechanism scans the input sequence and chooses relevant frames. We propose two methods to speed up this operation: limiting the scan to a subset of most promising frames and pooling over time the information contained in neighboring frames, thereby reducing source sequence length. Index Terms -- neural networks, L VCSR, attention, speech recognition, ASR 1. INTRODUCTION Deep neural networks have become popular acoustic models for state-of-the-art large vocabulary speech recognition systems (Hinton et al., 2012a). However, in these systems most of the other components, such as Hidden Markov Models (HMMs), Gaussian Mixture Models (GMMs) andn -gram language models, are the same as in their predecessors. These combinations of neural networks and statistical models are often referred to as hybrid systems.

In these notes, we wish to provide an introduction to the signature method, focusing on its basic theoretical properties and recent numerical applications. The notes are split into two parts. The first part focuses on the definition and fundamental properties of the signature of a path, or the path signature. We have aimed for a minimalistic approach, assuming only familiarity with classical real analysis and integration theory, and supplementing theory with straightforward examples. We have chosen to focus in detail on the principle properties of the signature which we believe are fundamental to understanding its role in applications. We also present an informal discussion on some of its deeper properties and briefly mention the role of the signature in rough paths theory, which we hope could serve as a light introduction to rough paths for the interested reader. The second part of these notes discusses practical applications of the path signature to the area of machine learning. The signature approach represents a non-parametric way for extraction of characteristic features from data. The data are converted into a multi-dimensional path by means of various embedding algorithms and then processed for computation of individual terms of the signature which summarise certain information contained in the data. The signature thus transforms raw data into a set of features which are used in machine learning tasks. We will review current progress in applications of signatures to machine learning problems.