Leveraging offline data is an attractive way to accelerate online sequential decision-making. However, it is crucial to account for latent states in users or environments in the offline data, and latent bandits form a compelling model for doing so. In this light, we design end-to-end latent bandit algorithms capable of handing uncountably many latent states. We focus on a linear latent contextual bandit — a linear bandit where each user has its own high-dimensional reward parameter in $\mathbb{R}^{d_A}$, but reward parameters across users lie in a low-rank latent subspace of dimension $d_K \ll d_A$. First, we provide an offline algorithm to learn this subspace with provable guarantees. We then present two online algorithms that utilize the output of this offline algorithm to accelerate online learning. The first enjoys $\tilde O(\min(d_A\sqrt{T}, d_K\sqrt{T}(1+\sqrt{d_AT/d_KN})))$ regret guarantees, so that the effective dimension is lower when the size $N$ of the offline dataset is larger. We prove a matching lower bound on regret, showing that our algorithm is minimax optimal. The second is a practical algorithm that enjoys only a slightly weaker guarantee, but is computationally efficient. We also establish the efficacy of our methods using experiments on both synthetic data and real-life movie recommendation data from MovieLens. Finally, we theoretically establish the generality of the latent bandit model by proving a de Finetti theorem for stateless decision processes.

Actor-critic algorithms have become a cornerstone in reinforcement learning (RL), leveraging the strengths of both policy-based and value-based methods. Despite recent progress in understanding their statistical efficiency, no existing work has successfully learned an $\epsilon$-optimal policy with a sample complexity of $O(1/\epsilon^2)$ trajectories with general function approximation when strategic exploration is necessary. We address this open problem by introducing a novel actor-critic algorithm that attains a sample-complexity of $O(dH^5 \log|\mathcal{A}|/\epsilon^2 + d H^4 \log|\mathcal{F}|/ \epsilon^2)$ trajectories, and accompanying $\sqrt{T}$ regret when the Bellman eluder dimension $d$ does not increase with $T$ at more than a $\log T$ rate. Here, $\mathcal{F}$ is the critic function class, and $\mathcal{A}$ is the action space. Our algorithm integrates optimism, off-policy critic estimation targeting the optimal Q-function, and rare-switching policy resets. We extend this to the setting of Hybrid RL, where we show that initializing the critic with offline data yields sample efficiency gains, and also provide a non-optimistic provably efficient actor-critic algorithm, addressing another open problem in the literature. Numerical experiments support our theoretical findings.

Gradient boosting is widely popular due to its flexibility and predictive accuracy. However, statistical inference and uncertainty quantification for gradient boosting remain challenging and under-explored. We propose a unified framework for statistical inference in gradient boosting regression. Our framework integrates dropout or parallel training with a recently proposed regularization procedure called Boulevard that allows for a central limit theorem (CLT) for boosting. With these enhancements, we surprisingly find that \textit{increasing} the dropout rate and the number of trees grown in parallel at each iteration substantially enhances signal recovery and overall performance. Our resulting algorithms enjoy similar CLTs, which we use to construct built-in confidence intervals, prediction intervals, and rigorous hypothesis tests for assessing variable importance in only $O(nd^2)$ time with the Nyström method. Numerical experiments verify the asymptotic normality and demonstrate that our algorithms perform well, do not require early stopping, interpolate between regularized boosting and random forests, and confirm the validity of their built-in statistical inference procedures.

Evaluating and optimizing policies in the presence of unobserved confounders is a problem of growing interest in offline reinforcement learning. Using conventional methods for offline RL in the presence of confounding can not only lead to poor decisions and poor policies, but also have disastrous effects in critical applications such as healthcare and education. We map out the landscape of offline policy evaluation for confounded MDPs, distinguishing assumptions on confounding based on whether they are memoryless and on their effect on the data-collection policies. We characterize settings where consistent value estimates are provably not achievable, and provide algorithms with guarantees to instead estimate lower bounds on the value. When consistent estimates are achievable, we provide algorithms for value estimation with sample complexity guarantees. We also present new algorithms for offline policy improvement and prove local convergence guarantees. Finally, we experimentally evaluate our algorithms on both a gridworld environment and a simulated healthcare setting of managing sepsis patients. In gridworld, our model-based method provides tighter lower bounds than existing methods, while in the sepsis simulator, we demonstrate the effectiveness of our method and investigate the importance of a clustering sub-routine.

Matrix sketching is a powerful tool for reducing the size of large data matrices. Yet there are fundamental limitations to this size reduction when we want to recover an accurate estimator for a task such as least square regression. We show that these limitations can be circumvented in the distributed setting by designing sketching methods that minimize the bias of the estimator, rather than its error. In particular, we give a sparse sketching method running in optimal space and current matrix multiplication time, which recovers a nearly-unbiased least squares estimator using two passes over the data. This leads to new communication-efficient distributed averaging algorithms for least squares and related tasks, which directly improve on several prior approaches. Our key novelty is a new bias analysis for sketched least squares, giving a sharp characterization of its dependence on the sketch sparsity. The techniques include new higher moment restricted Bai-Silverstein inequalities, which are of independent interest to the non-asymptotic analysis of deterministic equivalents for random matrices that arise from sketching.

Hybrid Reinforcement Learning (RL), where an agent learns from both an offline dataset and online explorations in an unknown environment, has garnered significant recent interest. A crucial question posed by Xie et al. (2022) is whether hybrid RL can improve upon the existing lower bounds established in purely offline and purely online RL without relying on the single-policy concentrability assumption. While Li et al. (2023) provided an affirmative answer to this question in the tabular PAC RL case, the question remains unsettled for both the regret-minimizing RL case and the non-tabular case. In this work, building upon recent advancements in offline RL and reward-agnostic exploration, we develop computationally efficient algorithms for both PAC and regret-minimizing RL with linear function approximation, without requiring concentrability on the entire state-action space. We demonstrate that these algorithms achieve sharper error or regret bounds that are no worse than, and can improve on, the optimal sample complexity in offline RL (the first algorithm, for PAC RL) and online RL (the second algorithm, for regret-minimizing RL) in linear Markov decision processes (MDPs), regardless of the quality of the behavior policy. To our knowledge, this work establishes the tightest theoretical guarantees currently available for hybrid RL in linear MDPs.

Public health decisions must be made about when and how to implement interventions to control an infectious disease epidemic. These decisions should be informed by data on the epidemic as well as current understanding about the transmission dynamics. Such decisions can be posed as statistical questions about scientifically motivated dynamic models. Thus, we encounter the methodological task of building credible, data-informed decisions based on stochastic, partially observed, nonlinear dynamic models. This necessitates addressing the tradeoff between biological fidelity and model simplicity, and the reality of misspecification for models at all levels of complexity. We assess current methodological approaches to these issues via a case study of the 2010-2019 cholera epidemic in Haiti. We consider three dynamic models developed by expert teams to advise on vaccination policies. We evaluate previous methods used for fitting these models, and we demonstrate modified data analysis strategies leading to improved statistical fit. Specifically, we present approaches for diagnosing model misspecification and the consequent development of improved models. Additionally, we demonstrate the utility of recent advances in likelihood maximization for high-dimensional nonlinear dynamic models, enabling likelihood-based inference for spatiotemporal incidence data using this class of models. Our workflow is reproducible and extendable, facilitating future investigations of this disease system.

We present an algorithm for learning mixtures of Markov chains and Markov decision processes (MDPs) from short unlabeled trajectories. Specifically, our method handles mixtures of Markov chains with optional control input by going through a multi-step process, involving (1) a subspace estimation step, (2) spectral clustering of trajectories using "pairwise distance estimators," along with refinement using the EM algorithm, (3) a model estimation step, and (4) a classification step for predicting labels of new trajectories. We provide end-to-end performance guarantees, where we only explicitly require the length of trajectories to be linear in the number of states and the number of trajectories to be linear in a mixing time parameter. Experimental results support these guarantees, where we attain 96.6% average accuracy on a mixture of two MDPs in gridworld, outperforming the EM algorithm with random initialization (73.2% average accuracy). We also significantly outperform the EM algorithm on real data from the LastFM song dataset.

Proactive inspection is essential for prediction and prevention of rolling stock component failures. The conventional process for inspecting bogies under trains presents significant challenges for inspectors who need to visually check the tight and cluttered environment. We propose a miniature multi-link climbing robot, called BogieBot, that can be deployed inside the undercarriage areas of trains and other large vehicles for inspection and maintenance purposes. BogieBot can carry a visual sensor or manipulator on its main body. The novel compact design utilises six identical couple joints and two mechanically switchable magnetic grippers that together, empower multi-modal climbing and manipulation. The proposed mechanism is kinematically redundant, allowing the robot to perform self-motions in a tight space and manoeuvre around obstacles. The mechanism design and various analyses on the forward and inverse kinematic, work-space, and self-motions of BogieBot are presented. The robot is demonstrated to perform challenging navigation tasks in different scenarios involving simulated complex environments.

Micro-facial expressions are spontaneous, involuntary movements of the face when a person experiences an emotion but attempts to hide their facial expression, most likely in a high-stakes environment. Recently, research in this field has grown in popularity, however publicly available datasets of micro-expressions have limitations due to the difficulty of naturally inducing spontaneous micro-expressions. Other issues include lighting, low resolution and low participant diversity. We present a newly developed spontaneous micro-facial movement dataset with diverse participants and coded using the Facial Action Coding System. The experimental protocol addresses the limitations of previous datasets, including eliciting emotional responses from stimuli tailored to each participant. Dataset evaluation was completed by running preliminary experiments to classify micro-movements from non-movements. Results were obtained using a selection of spatio-temporal descriptors and machine learning. We further evaluate the dataset on emerging methods of feature difference analysis and propose an Adaptive Baseline Threshold that uses individualised neutral expression to improve the performance of micro-movement detection. In contrast to machine learning approaches, we outperform the state of the art with a recall of 0.91. The outcomes show the dataset can become a new standard for micro-movement data, with future work expanding on data representation and analysis.

Annotation of data is fundamental for training any modern facial tracking system. Methods, such as deep and machine learning, require large amounts of pre-annotated data to produce results found in many state-of-the-art systems. 2- Dimensional (images) annotations rely on the texture information of the face to annotate the features. However, in 3-Dimensions (models), the task becomes more complex. In 3D, the conventional 2D approaches are ineffective as facial landmarks are difficult to accurately identify without texture information. There has been little research into the accuracy of methods for annotating 3D facial models. This paper proposes a method for annotating 3D models which uses texture information by aligning a model to a 2D image to compare its accuracy and throughput to the conventional methods of 3D annotation. For evaluation, 16 nonexpert volunteers were recruited and instructed to annotate three models using both the proposed method and the conventional method. The resultant annotations were compared to ground truth data generated by an experienced annotator. The results demonstrate significant improvement in the throughput of the proposed annotation method compared to the conventional approach without significant differences in accuracy. The proposed method also highlights that the conventional method does not successfully identify all the facial landmarks. The proposed method will be made freely available to use online.

An experimental comparison of interaction in the real world and a CAVE virtual environment was carried out, varying interaction with and without virtual hands and comparing two manipulation tasks. The double-handed task was possible in the real world but nearly impossible in the VE, leading to changed behavior. The single-handed task showed more errors in the VE but few behavioral differences. Users encountered more errors in the CAVE condition without the virtual hand than with it, and few errors in the real world. Visual feedback caused many usability problems in both tasks. The implications for VE usability and virtual prototyping are discussed.

Virtual environment technology is now beginning to be recognised as a powerful design tool in industrial sectors such as Manufacturing, Process Engineering, Construction, Automotive and Aerospace industries. It offers the ability to visualise a design from different viewpoints by engineers from different design perspectives providing a powerful design analysis tool for supporting concurrent engineering philosophy. A common weakness of the current commercial virtual environments is the lack of efficient geometric constraint management facilities such as run-time constraint detection and the maintenance of constraint consistencies for supporting accurate part positioning and constrained 3D manipulations. The environments also need to be designed to support the user as they are completing their task. This paper describes the software architecture of a constraint-based virtual environment that supports interactive assembly of component parts, embedded within a task based environment that supports contextual help and allows for the structure of tasks to be easily altered for rapid prototyping.