We consider the joint design and control of discrete-time stochastic dynamical systems over a finite time horizon. We formulate the problem as a multi-step optimization problem under uncertainty seeking to identify a system design and a control policy that jointly maximize the expected sum of rewards collected over the time horizon considered. The transition function, the reward function and the policy are all parametrized, assumed known and differentiable with respect to their parameters. We then introduce a deep reinforcement learning algorithm combining policy gradient methods with model-based optimization techniques to solve this problem. In essence, our algorithm iteratively approximates the gradient of the expected return via Monte-Carlo sampling and automatic differentiation and takes projected gradient ascent steps in the space of environment and policy parameters. This algorithm is referred to as Direct Environment and Policy Search (DEPS). We assess the performance of our algorithm in three environments concerned with the design and control of a mass-spring-damper system, a small-scale off-grid power system and a drone, respectively. In addition, our algorithm is benchmarked against a state-of-the-art deep reinforcement learning algorithm used to tackle joint design and control problems. We show that DEPS performs at least as well or better in all three environments, consistently yielding solutions with higher returns in fewer iterations. Finally, solutions produced by our algorithm are also compared with solutions produced by an algorithm that does not jointly optimize environment and policy parameters, highlighting the fact that higher returns can be achieved when joint optimization is performed.

We suggest a simple practical method to combine the human and artificial intelligence to both learn best investment practices of fund managers, and provide recommendations to improve them. Our approach is based on a combination of Inverse Reinforcement Learning (IRL) and RL. First, the IRL component learns the intent of fund managers as suggested by their trading history, and recovers their implied reward function. At the second step, this reward function is used by a direct RL algorithm to optimize asset allocation decisions. We show that our method is able to improve over the performance of individual fund managers.

Estimating value functions is a core component of reinforcement learning algorithms. Temporal difference (TD) learning algorithms use bootstrapping, i.e. they update the value function toward a learning target using value estimates at subsequent time-steps. Alternatively, the value function can be updated toward a learning target constructed by separately predicting successor features (SF)--a policy-dependent model--and linearly combining them with instantaneous rewards. We focus on bootstrapping targets used when estimating value functions, and propose a new backup target, the $\eta$-return mixture, which implicitly combines value-predictive knowledge (used by TD methods) with (successor) feature-predictive knowledge--with a parameter $\eta$ capturing how much to rely on each. We illustrate that incorporating predictive knowledge through an $\eta\gamma$-discounted SF model makes more efficient use of sampled experience, compared to either extreme, i.e. bootstrapping entirely on the value function estimate, or bootstrapping on the product of separately estimated successor features and instantaneous reward models. We empirically show this approach leads to faster policy evaluation and better control performance, for tabular and nonlinear function approximations, indicating scalability and generality.

A key challenge in the study of multiagent cooperation is the need for individual agents not only to cooperate effectively, but to decide with whom to cooperate. This is particularly critical in situations when other agents have hidden, possibly misaligned motivations and goals. Social deduction games offer an avenue to study how individuals might learn to synthesize potentially unreliable information about others, and elucidate their true motivations. In this work, we present Hidden Agenda, a two-team social deduction game that provides a 2D environment for studying learning agents in scenarios of unknown team alignment. The environment admits a rich set of strategies for both teams. Reinforcement learning agents trained in Hidden Agenda show that agents can learn a variety of behaviors, including partnering and voting without need for communication in natural language.

Evaluating the real-world performance of network protocols is challenging. Randomized control trials (RCT) are expensive and inaccessible to most researchers, while expert-designed simulators fail to capture complex behaviors in real networks. We present CausalSim, a data-driven simulator for network protocols that addresses this challenge. Learning network behavior from observational data is complicated due to the bias introduced by the protocols used during data collection. CausalSim uses traces from an initial RCT under a set of protocols to learn a causal network model, effectively removing the biases present in the data. Using this model, CausalSim can then simulate any protocol over the same traces (i.e., for counterfactual predictions). Key to CausalSim is the novel use of adversarial neural network training that exploits distributional invariances that are present due to the training data coming from an RCT. Our extensive evaluation of CausalSim on both real and synthetic datasets and two use cases, including more than nine months of real data from the Puffer video streaming system, shows that it provides accurate counterfactual predictions, reducing prediction error by 44% and 53% on average compared to expert-designed and standard supervised learning baselines.

In model-free deep reinforcement learning (RL) algorithms, using noisy value estimates to supervise policy evaluation and optimization is detrimental to the sample efficiency. As this noise is heteroscedastic, its effects can be mitigated using uncertainty-based weights in the optimization process. Previous methods rely on sampled ensembles, which do not capture all aspects of uncertainty. We provide a systematic analysis of the sources of uncertainty in the noisy supervision that occurs in RL, and introduce inverse-variance RL, a Bayesian framework which combines probabilistic ensembles and Batch Inverse Variance weighting. We propose a method whereby two complementary uncertainty estimation methods account for both the Q-value and the environment stochasticity to better mitigate the negative impacts of noisy supervision. Our results show significant improvement in terms of sample efficiency on discrete and continuous control tasks.

In recent years, voice-based virtual assistants such as Google Assistant and Amazon Alexa have grown popular. This has presented both potential and challenges for natural language understanding (NLU) systems. These devices' production systems are often trained by supervised learning and rely significantly on annotated data. But, data annotation is costly and time-consuming. Furthermore, model updates using offline supervised learning can take long and miss trending requests.

Deep reinforcement learning has gathered much attention recently. Impressive results were achieved in activities as diverse as autonomous driving, game playing, molecular recombination, and robotics. In all these fields, computer programs have taught themselves to solve difficult problems. They have learned to fly model helicopters and perform aerobatic manoeuvers such as loops and rolls. In some applications they have even become better than the best humans, such as in Atari, Go, poker and StarCraft. The way in which deep reinforcement learning explores complex environments reminds us of how children learn, by playfully trying out things, getting feedback, and trying again. The computer seems to truly possess aspects of human learning; this goes to the heart of the dream of artificial intelligence. The successes in research have not gone unnoticed by educators, and universities have started to offer courses on the subject. The aim of this book is to provide a comprehensive overview of the field of deep reinforcement learning. The book is written for graduate students of artificial intelligence, and for researchers and practitioners who wish to better understand deep reinforcement learning methods and their challenges. We assume an undergraduate-level of understanding of computer science and artificial intelligence; the programming language of this book is Python. We describe the foundations, the algorithms and the applications of deep reinforcement learning. We cover the established model-free and model-based methods that form the basis of the field. Developments go quickly, and we also cover advanced topics: deep multi-agent reinforcement learning, deep hierarchical reinforcement learning, and deep meta learning.

Programming intelligent control strategies for complex robot systems is a challenging task. Reinforcement learning (RL) promises the automated synthesis of control strategies through a data-driven approach instead of explicitly designing hand-crafted solutions through expert knowledge. In recent years, the field of RL has witnessed outstanding successes and raised a surge of interest in the control of dynamical systems through such a trial-and-error paradigm. The combination of RL and deep learning methods excel at problems that can be quickly simulated like robotics [13, 22] or video games [18, 29] and in domains where the exact model is known but long-horizon planning is not computationally tractable, e.g. board games like Go and Chess [26]. Despite the significant advances in recent years, the applicability of RL algorithms is still limited when the data at test time differ from those seen during training [10]. Since many real-world systems cannot afford to learn policies from scratch due to the expense of data, simulations are the preferred approach to build data-driven control policies in the RL community. In fact, a gap between the simulation and the real world still persists because the modeling of all effects either requires in-depth expert knowledge or is simply not desirable, e.g.

The ability to recover from an unexpected external perturbation is a fundamental motor skill in bipedal locomotion. An effective response includes the ability to not just recover balance and maintain stability but also to fall in a safe manner when balance recovery is physically infeasible. For robots associated with bipedal locomotion, such as humanoid robots and assistive robotic devices that aid humans in walking, designing controllers which can provide this stability and safety can prevent damage to robots or prevent injury related medical costs. This is a challenging task because it involves generating highly dynamic motion for a high-dimensional, non-linear and under-actuated system with contacts. Despite prior advancements in using model-based and optimization methods, challenges such as requirement of extensive domain knowledge, relatively large computational time and limited robustness to changes in dynamics still make this an open problem. In this thesis, to address these issues we develop learning-based algorithms capable of synthesizing push recovery control policies for two different kinds of robots : Humanoid robots and assistive robotic devices that assist in bipedal locomotion. Our work can be branched into two closely related directions : 1) Learning safe falling and fall prevention strategies for humanoid robots and 2) Learning fall prevention strategies for humans using a robotic assistive devices. To achieve this, we introduce a set of Deep Reinforcement Learning (DRL) algorithms to learn control policies that improve safety while using these robots.