With the advent of large datasets, offline reinforcement learning (RL) is a promising framework for learning good decision-making policies without the need to interact with the real environment. However, offline RL requires the dataset to be reward-annotated, which presents practical challenges when reward engineering is difficult or when obtaining reward annotations is labor-intensive. In this paper, we introduce Optimal Transport Reward labeling (OTR), an algorithm that assigns rewards to offline trajectories, with a few high-quality demonstrations. OTR's key idea is to use optimal transport to compute an optimal alignment between an unlabeled trajectory in the dataset and an expert demonstration to obtain a similarity measure that can be interpreted as a reward, which can then be used by an offline RL algorithm to learn the policy. OTR is easy to implement and computationally efficient. On D4RL benchmarks, we show that OTR with a single demonstration can consistently match the performance of offline RL with ground-truth rewards.

Within a robotic context, we merge the techniques of passivity-based control (PBC) and reinforcement learning (RL) with the goal of eliminating some of their reciprocal weaknesses, as well as inducing novel promising features in the resulting framework. We frame our contribution in a scenario where PBC is implemented by means of virtual energy tanks, a control technique developed to achieve closed-loop passivity for any arbitrary control input. Albeit the latter result is heavily used, we discuss why its practical application at its current stage remains rather limited, which makes contact with the highly debated claim that passivity-based techniques are associated with a loss of performance. The use of RL allows us to learn a control policy that can be passivized using the energy tank architecture, combining the versatility of learning approaches and the system theoretic properties which can be inferred due to the energy tanks. Simulations show the validity of the approach, as well as novel interesting research directions in energy-aware robotics.

The aim of Inverse Reinforcement Learning (IRL) is to infer a reward function $R$ from a policy $\pi$. To do this, we need a model of how $\pi$ relates to $R$. In the current literature, the most common models are optimality, Boltzmann rationality, and causal entropy maximisation. One of the primary motivations behind IRL is to infer human preferences from human behaviour. However, the true relationship between human preferences and human behaviour is much more complex than any of the models currently used in IRL. This means that they are misspecified, which raises the worry that they might lead to unsound inferences if applied to real-world data. In this paper, we provide a mathematical analysis of how robust different IRL models are to misspecification, and answer precisely how the demonstrator policy may differ from each of the standard models before that model leads to faulty inferences about the reward function $R$. We also introduce a framework for reasoning about misspecification in IRL, together with formal tools that can be used to easily derive the misspecification robustness of new IRL models.

Due to their superior energy efficiency, blimps may replace quadcopters for long-duration aerial tasks. However, designing a controller for blimps to handle complex dynamics, modeling errors, and disturbances remains an unsolved challenge. One recent work combines reinforcement learning (RL) and a PID controller to address this challenge and demonstrates its effectiveness in real-world experiments. In the current work, we build on that using an H-infinity robust controller to expand the stability margin and improve the RL agent's performance. Empirical analysis of different mixing methods reveals that the resulting H-infinity-RL controller outperforms the prior PID-RL combination and can handle more complex tasks involving intensive thrust vectoring. We provide our code as open-source at https://github.com/robot-perception-group/robust_deep_residual_blimp.

We consider the problem of controlling an unknown stochastic linear system with quadratic costs - called the adaptive LQ control problem. We re-examine an approach called ''Reward Biased Maximum Likelihood Estimate'' (RBMLE) that was proposed more than forty years ago, and which predates the ''Upper Confidence Bound'' (UCB) method as well as the definition of ''regret'' for bandit problems. It simply added a term favoring parameters with larger rewards to the criterion for parameter estimation. We show how the RBMLE and UCB methods can be reconciled, and thereby propose an Augmented RBMLE-UCB algorithm that combines the penalty of the RBMLE method with the constraints of the UCB method, uniting the two approaches to optimism in the face of uncertainty. We establish that theoretically, this method retains $\Tilde{\mathcal{O}}(\sqrt{T})$ regret, the best-known so far. We further compare the empirical performance of the proposed Augmented RBMLE-UCB and the standard RBMLE (without the augmentation) with UCB, Thompson Sampling, Input Perturbation, Randomized Certainty Equivalence and StabL on many real-world examples including flight control of Boeing 747 and Unmanned Aerial Vehicle. We perform extensive simulation studies showing that the Augmented RBMLE consistently outperforms UCB, Thompson Sampling and StabL by a huge margin, while it is marginally better than Input Perturbation and moderately better than Randomized Certainty Equivalence.

In this paper, we revisit endless online level generation with the recently proposed experience-driven procedural content generation via reinforcement learning (EDRL) framework. Inspired by an observation that EDRL tends to generate recurrent patterns, we formulate a notion of state space closure which makes any stochastic state appeared possibly in an infinite-horizon online generation process can be found within a finite-horizon. Through theoretical analysis, we find that even though state space closure arises a concern about diversity, it generalises EDRL trained with a finite-horizon to the infinite-horizon scenario without deterioration of content quality. Moreover, we verify the quality and the diversity of contents generated by EDRL via empirical studies, on the widely used Super Mario Bros. benchmark. Experimental results reveal that the diversity of levels generated by EDRL is limited due to the state space closure, whereas their quality does not deteriorate in a horizon which is longer than the one specified in the training. Concluding our outcomes and analysis, future work on endless online level generation via reinforcement learning should address the issue of diversity while assuring the occurrence of state space closure and quality.

Socially Assistive Robots (SARs) have shown promising potential in therapeutic scenarios as decision-making instructors or motivational companions. In human-human therapy, experts often communicate the thought process behind the decisions they make to promote transparency and build trust. As research aims to incorporate more complex decision-making models into these robots to drive better interaction, the ability for the SAR to explain its decisions becomes an increasing challenge. We present the latest examples of complex SAR decision-makers. We argue that, based on the importance of transparent communication in human-human therapy, SARs should incorporate such components into their design. To stimulate discussion around this topic, we present a set of design considerations for researchers.

Interest in reinforcement learning (RL, Sutton & Barto 2018) has increased dramatically in recent years due in part to a number of high-profile successes in games (Mnih et al. 2013, 2015), autonomous driving (Sallab et al. 2017), and precision medicine (Tsiatis et al. 2019). However, despite theoretical and computational advances, real-world application of RL remains difficult. A primary challenge is dealing with high-dimensional state representations. Such representations occur naturally in systems with high-dimensional measurements, like images or audio, but can also occur when the system state is constructed by concatenating a series of measurements over a contiguous block of time. A high-dimensional state-- when a more parsimonious one would suffice--dilutes the efficiency of learning algorithms and makes the estimated optimal policy harder to interpret. Thus, methods for removing uninformative or redundant variables from the state are of tremendous practical value. We develop a general variable selection algorithm for offline RL, which aims to learn an optimal policy using only logged data, i.e., without any additional online interaction. Our contributions can be summarized as follows: (i) we formally define a minimal sufficient state for an MDP and argue that it is an appropriate target by which to design and evaluate variable selection methods in RL; (ii) we show that naïve variable selection methods based on the state or reward alone need not recover the minimal sufficient state; (iii) we propose a novel sequential knockoffs (SEEK) algorithm that applies with general black-box learning methods, and, under a β-mixing condition, consistently recovers the minimal sufficient state, and controls the false discovery rate (FDR, the ratio of the number of selected irrelevant variables to the number of selected variables); and (iv) we develop a novel algorithm to estimate the β-mixing coefficients of an MDP. The algorithm in (iv) is important in its own right as it applies to a number of applications beyond RL (McDonald et al. 2015).

This study proposes a safe and sample-efficient reinforcement learning (RL) framework to address two major challenges in developing applicable RL algorithms: satisfying safety constraints and efficiently learning with limited samples. To guarantee safety in real-world complex environments, we use the safe set algorithm (SSA) to monitor and modify the nominal controls, and evaluate SSA+RL in a clustered dynamic environment which is challenging to be solved by existing RL algorithms. However, the SSA+RL framework is usually not sample-efficient especially in reward-sparse environments, which has not been addressed in previous safe RL works. To improve the learning efficiency, we propose three techniques: (1) avoiding behaving overly conservative by adapting the SSA; (2) encouraging safe exploration using random network distillation with safety constraints; (3) improving policy convergence by treating SSA as expert demonstrations and directly learn from that. The experimental results show that our framework can achieve better safety performance compare to other safe RL methods during training and solve the task with substantially fewer episodes. Project website: https://hychen-naza.github.io/projects/Safe_RL/.

In counter-adversarial systems, to infer the strategy of an intelligent adversarial agent, the defender agent needs to cognitively sense the information that the adversary has gathered about the latter. Prior works on the problem employ linear Gaussian state-space models and solve this inverse cognition problem by designing inverse stochastic filters. However, in practice, counter-adversarial systems are generally highly nonlinear. In this paper, we address this scenario by formulating inverse cognition as a nonlinear Gaussian state-space model, wherein the adversary employs an unscented Kalman filter (UKF) to estimate the defender's state with reduced linearization errors. To estimate the adversary's estimate of the defender, we propose and develop an inverse UKF (IUKF) system. We then derive theoretical guarantees for the stochastic stability of IUKF in the mean-squared boundedness sense. Numerical experiments for multiple practical applications show that the estimation error of IUKF converges and closely follows the recursive Cram\'{e}r-Rao lower bound.