Stochastic Approximation (SA) is a popular approach for solving fixed point equations where the information is corrupted by noise. In this paper, we consider an SA involving a contraction mapping with respect to an arbitrary norm, and show its finite-sample bound for using either constant or diminishing step sizes. The idea is to construct a smooth Lyapunov function using the generalized Moreau envelope, and show that the iterates of SA are contracting in expectation with respect to that Lyapunov function. The result is applicable to various Reinforcement Learning (RL) algorithms. In particular, we use it to establish the first-known convergence rate of the V-trace algorithm for the off-policy TD-Learning [15], and improve the existing bound for the tabular $Q$-Learning algorithm. Further, for these two applications, our construction of the Lyapunov functions results in only a logarithmic dependence of the convergence bound on the state-space dimension.

With the development of deep representation learning, the domain of reinforcement learning (RL) has become a powerful learning framework now capable of learning complex policies in high dimensional environments. This review summarises deep reinforcement learning (DRL) algorithms, provides a taxonomy of automated driving tasks where (D)RL methods have been employed, highlights the key challenges algorithmically as well as in terms of deployment of real world autonomous driving agents, the role of simulators in training agents, and finally methods to evaluate, test and robustifying existing solutions in RL and imitation learning.

Automated driving in urban settings is challenging chiefly due to the indeterministic nature of the human participants of the traffic. These behaviors are difficult to model, and conventional, rule-based Automated Driving Systems (ADSs) tend to fail when they face unmodeled dynamics. On the other hand, the more recent, end-to-end Deep Reinforcement Learning (DRL) based ADSs have shown promising results. However, pure learning-based approaches lack the hard-coded safety measures of model-based methods. Here we propose a hybrid approach that integrates a model-based path planner into a vision based DRL framework to alleviate the shortcomings of both worlds. In summary, the DRL agent learns to overrule the model-based planner's decisions if it predicts that better future rewards can be obtained while doing so, e.g., avoiding an accident. Otherwise, the DRL agent tends to follow the model-based planner as close as possible. This logic is learned, i.e., no switching model is designed here. The agent learns this by considering two penalties: the penalty of straying away from the model-based path planner and the penalty of having a collision. The latter has precedence over the former, i.e., the penalty is greater. Therefore, after training, the agent learns to follow the model-based planner when it is safe to do so, otherwise, it gets penalized. However, it also learns to sacrifice positive rewards for following the model-based planner to avoid a potential big negative penalty for making a collision in the future. Experimental results show that the proposed method can plan its path and navigate while avoiding obstacles between randomly chosen origin-destination points in CARLA, a dynamic urban simulation environment. Our code is open-source and available online.

In adversarial imitation learning, a discriminator is trained to differentiate agent episodes from expert demonstrations representing the desired behavior. However, as the trained policy learns to be more successful, the negative examples (the ones produced by the agent) become increasingly similar to expert ones. Despite the fact that the task is successfully accomplished in some of the agent's trajectories, the discriminator is trained to output low values for them. We hypothesize that this inconsistent training signal for the discriminator can impede its learning, and consequently leads to worse overall performance of the agent. We show experimental evidence for this hypothesis and that the 'False Negatives' (i.e. successful agent episodes) significantly hinder adversarial imitation learning, which is the first contribution of this paper. Then, we propose a method to alleviate the impact of false negatives and test it on the BabyAI environment. This method consistently improves sample efficiency over the baselines by at least an order of magnitude.

Welcome to Deep Reinforcement Learning 2.0! In this course, we will learn and implement a new incredibly smart AI model, called the Twin-Delayed DDPG, which combines state of the art techniques in Artificial Intelligence including continuous Double Deep Q-Learning, Policy Gradient, and Actor Critic. The model is so strong that for the first time in our courses, we are able to solve the most challenging virtual AI applications (training an ant/spider and a half humanoid to walk and run across a field). In this part we will study all the fundamentals of Artificial Intelligence which will allow you to understand and master the AI of this course. These include Q-Learning, Deep Q-Learning, Policy Gradient, Actor-Critic and more.

Off-policy ensemble reinforcement learning (RL) methods have demonstrated impressive results across a range of RL benchmark tasks. Recent works suggest that directly imitating experts' policies in a supervised manner before or during the course of training enables faster policy improvement for an RL agent. Motivated by these recent insights, we propose Periodic Intra-Ensemble Knowledge Distillation (PIEKD). PIEKD is a learning framework that uses an ensemble of policies to act in the environment while periodically sharing knowledge amongst policies in the ensemble through knowledge distillation. Our experiments demonstrate that PIEKD improves upon a state-of-the-art RL method in sample efficiency on several challenging MuJoCo benchmark tasks. Additionally, we perform ablation studies to better understand PIEKD.

This paper reports on the open source project CLAI (Command Line AI), aimed at bringing the power of AI to the command line interface. The platform sets up the CLI as a new environment for AI researchers to conquer by surfacing the command line as a generic environment that researchers can interface to using a simple sense-act API much like the traditional AI agent architecture. In this paper, we discuss the design and implementation of the platform in detail, through illustrative use cases of new end user interaction patterns enabled by this design, and through quantitative evaluation of the system footprint of a CLAI-enabled terminal. We also report on some early user feedback on its features from an internal survey.

We study locally differentially private algorithms for reinforcement learning to obtain a robust policy that performs well across distributed private environments. Our algorithm protects the information of local agents' models from being exploited by adversarial reverse engineering. Since a local policy is strongly being affected by the individual environment, the output of the agent may release the private information unconsciously. In our proposed algorithm, local agents update the model in their environments and report noisy gradients designed to satisfy local differential privacy (LDP) that gives a rigorous local privacy guarantee. By utilizing a set of reported noisy gradients, a central aggregator updates its model and delivers it to different local agents. In our empirical evaluation, we demonstrate how our method performs well under LDP. To the best of our knowledge, this is the first work that actualizes distributed reinforcement learning under LDP. This work enables us to obtain a robust agent that performs well across distributed private environments.

It is difficult to be able to imitate well in unknown states from a small amount of expert data and sampling data. Supervised learning methods such as Behavioral Cloning do not require sampling data, but usually suffer from distribution shift. The methods based on reinforcement learning, such as inverse reinforcement learning and generative adversarial imitation learning (GAIL), can learn from only a few expert data. However, they often need to interact with the environment. Soft Q imitation learning addressed the problems, and it was shown that it could learn efficiently by combining Behavioral Cloning and soft Q-learning with constant rewards. In order to make this algorithm more robust to distribution shift, we propose Discriminator Soft Actor Critic (DSAC). It uses a reward function based on adversarial inverse reinforcement learning instead of constant rewards. We evaluated it on PyBullet environments with only four expert trajectories.

We present a novel negotiation model that allows an agent to learn how to negotiate during concurrent bilateral negotiations in unknown and dynamic e-markets. The agent uses an actor-critic architecture with model-free reinforcement learning to learn a strategy expressed as a deep neural network. We pre-train the strategy by supervision from synthetic market data, thereby decreasing the exploration time required for learning during negotiation. As a result, we can build automated agents for concurrent negotiations that can adapt to different e-market settings without the need to be pre-programmed. Our experimental evaluation shows that our deep reinforcement learning based agents outperform two existing well-known negotiation strategies in one-to-many concurrent bilateral negotiations for a range of e-market settings.