Q-learning is a model-free reinforcement learning algorithm to learn a policy telling an agent what action to take under what circumstances. It does not require a model of the environment, and it can handle problems with stochastic transitions and rewards, without requiring adaptations. For any finite Markov decision process (FMDP), Q-learning finds an optimal policy in the sense of maximizing the expected value of the total reward over any and all successive steps, starting from the current state. Q-learning can identify an optimal action-selection policy for any given FMDP, given infinite exploration time and a partly-random policy. "Q" names the function that returns the reward used to provide the reinforcement and can be said to stand for the "quality" of an action taken in a given state.

Model-based reinforcement learning (RL) is appealing because (i) it enables planning and thus more strategic exploration, and (ii) by decoupling dynamics from rewards, it enables fast transfer to new reward functions. However, learning an accurate Markov Decision Process (MDP) over high-dimensional states (e.g., raw pixels) is extremely challenging because it requires function approximation, which leads to compounding errors. Instead, to avoid compounding errors, we propose learning an abstract MDP over abstract states: low-dimensional coarse representations of the state (e.g., capturing agent position, ignoring other objects). We assume access to an abstraction function that maps the concrete states to abstract states. In our approach, we construct an abstract MDP, which grows through strategic exploration via planning. Similar to hierarchical RL approaches, the abstract actions of the abstract MDP are backed by learned subpolicies that navigate between abstract states. Our approach achieves strong results on three of the hardest Arcade Learning Environment games (Montezuma's Revenge, Pitfall!, and Private Eye), including superhuman performance on Pitfall! without demonstrations. After training on one task, we can reuse the learned abstract MDP for new reward functions, achieving higher reward in 1000x fewer samples than model-free methods trained from scratch.

Autonomous and semi-autonomous systems for safety-critical applications require rigorous testing before deployment. Due to the complexity of these systems, formal verification may be impossible and real-world testing may be dangerous during development. Therefore, simulation-based techniques have been developed that treat the system under test as a black box during testing. Safety validation tasks include finding disturbances to the system that cause it to fail (falsification), finding the most-likely failure, and estimating the probability that the system fails. Motivated by the prevalence of safety-critical artificial intelligence, this work provides a survey of state-of-the-art safety validation techniques with a focus on applied algorithms and their modifications for the safety validation problem. We present and discuss algorithms in the domains of optimization, path planning, reinforcement learning, and importance sampling. Problem decomposition techniques are presented to help scale algorithms to large state spaces, and a brief overview of safety-critical applications is given, including autonomous vehicles and aircraft collision avoidance systems. Finally, we present a survey of existing academic and commercially available safety validation tools.

The field of reinforcement learning can be split into model-based and model-free methods. Here, we unify these approaches by casting model-free policy optimisation as amortised variational inference, and model-based planning as iterative variational inference, within a `control as hybrid inference' (CHI) framework. We present an implementation of CHI which naturally mediates the balance between iterative and amortised inference. Using a didactic experiment, we demonstrate that the proposed algorithm operates in a model-based manner at the onset of learning, before converging to a model-free algorithm once sufficient data have been collected. We verify the scalability of our algorithm on a continuous control benchmark, demonstrating that it outperforms strong model-free and model-based baselines. CHI thus provides a principled framework for harnessing the sample efficiency of model-based planning while retaining the asymptotic performance of model-free policy optimisation.

Conventional simulations on multi-exit indoor evacuation focus primarily on how to determine a reasonable exit based on numerous factors in a changing environment. Results commonly include some congested and other under-utilized exits, especially with massive pedestrians. We propose a multi-exit evacuation simulation based on Deep Reinforcement Learning (DRL), referred to as the MultiExit-DRL, which involves in a Deep Neural Network (DNN) framework to facilitate state-to-action mapping. The DNN framework applies Rainbow Deep Q-Network (DQN), a DRL algorithm that integrates several advanced DQN methods, to improve data utilization and algorithm stability, and further divides the action space into eight isometric directions for possible pedestrian choices. We compare MultiExit-DRL with two conventional multi-exit evacuation simulation models in three separate scenarios: 1) varying pedestrian distribution ratios, 2) varying exit width ratios, and 3) varying open schedules for an exit. The results show that MultiExit-DRL presents great learning efficiency while reducing the total number of evacuation frames in all designed experiments. In addition, the integration of DRL allows pedestrians to explore other potential exits and helps determine optimal directions, leading to the high efficiency of exit utilization.

Deterministic Policy Gradient (DPG) removes a level of randomness from standard randomized-action Policy Gradient (PG), and demonstrates substantial empirical success for tackling complex dynamic problems involving Markov decision processes. At the same time, though, DPG loses its ability to learn in a model-free (i.e., actor-only) fashion, frequently necessitating the use of critics in order to obtain consistent estimates of the associated policy-reward gradient. In this work, we introduce Zeroth-order Deterministic Policy Gradient (ZDPG), which approximates policy-reward gradients via two-point stochastic evaluations of the $Q$-function, constructed by properly designed low-dimensional action-space perturbations. Exploiting the idea of random horizon rollouts for obtaining unbiased estimates of the $Q$-function, ZDPG lifts the dependence on critics and restores true model-free policy learning, while enjoying built-in and provable algorithmic stability. Additionally, we present new finite sample complexity bounds for ZDPG, which improve upon existing results by up to two orders of magnitude. Our findings are supported by several numerical experiments, which showcase the effectiveness of ZDPG in a practical setting, and its advantages over both PG and Baseline PG.

We introduce an approach for understanding finite-state machine (FSM) representations of recurrent policy networks. Recent work focused on minimizing FSMs to gain high-level insight, however, minimization can obscure a deeper understanding by merging states that are semantically distinct. Conversely, our approach starts with an unminimized machine and applies more-interpretable reductions that preserve the key decision points of the policy. We also contribute a saliency tool to attain a deeper understanding of the role of observations in the decisions. Our case studies on policies from 7 Atari games and 3 control benchmarks demonstrate that the approach can reveal insights that have not been noticed in prior work.

Deep Learning methods are replacing traditional software methods in solving real-world problems. Cheap and easily available computational power combined with labeled big datasets enabled deep learning algorithms to show their full potential. AlexNet paper(2012; Krizhevsky et al.[9]) showed feeding sufficient data into deep neural networks successfully learned to extract representations better than handcrafted features which let the start an era known as the rise of Deep Learning. Their great success in solving otherwise hard engineering problems such as object detection, voice recognition, chatbots, robotic manipulation and autonomous systems shown they can be applied to various fields thanks to their generalisation capability.[16] Path Planning(Motion Planning) is defined as computing a continuous path from starting position S to destination position D while avoiding any known obstacles in the way.[20] Whether it is in 2D or 3D geometry, any robotic system then will able to follow the computed path to reach it's destination. Real World robotic systems tend to use more explainable and reproducible algorithms based on interval based search (A star or Dijkstra) or sampling-based algorithms. We wanted to show a reward based algorithm that depends on Markov Decision Process(MDP) by trying to maximize cumulative future rewards can also complete long term path planning tasks. Advantage of using this option will allow autonomous robot(in our case simulated quadrotor) to create paths in non holonomic constraints which is something current methods fails to achieve.[1][17]

This paper serves as an introduction and overview of the potentially useful models and methodologies from artificial intelligence (AI) into the field of transportation engineering for autonomous vehicle (AV) control in the era of mixed autonomy. We will discuss state-of-the-art applications of AI-guided methods, identify opportunities and obstacles, raise open questions, and help suggest the building blocks and areas where AI could play a role in mixed autonomy. We divide the stage of autonomous vehicle (AV) deployment into four phases: the pure HVs, the HV-dominated, the AVdominated, and the pure AVs. This paper is primarily focused on the latter three phases. It is the first-of-its-kind survey paper to comprehensively review literature in both transportation engineering and AI for mixed traffic modeling. Models used for each phase are summarized, encompassing game theory, deep (reinforcement) learning, and imitation learning. While reviewing the methodologies, we primarily focus on the following research questions: (1) What scalable driving policies are to control a large number of AVs in mixed traffic comprised of human drivers and uncontrollable AVs? (2) How do we estimate human driver behaviors? (3) How should the driving behavior of uncontrollable AVs be modeled in the environment? (4) How are the interactions between human drivers and autonomous vehicles characterized? Hopefully this paper will not only inspire our transportation community to rethink the conventional models that are developed in the data-shortage era, but also reach out to other disciplines, in particular robotics and machine learning, to join forces towards creating a safe and efficient mixed traffic ecosystem.

Visualization tools for supervised learning have Visualization systems at their core consist of two components: allowed users to interpret, introspect, and gain representation and interaction. Though these may intuition for the successes and failures of their appear to be disparate, it is hard to discount the influence models. While reinforcement learning practitioners that each has on each other. The tools we use for representation ask many of the same questions, existing tools affect how we interact with the system, and our are not applicable to the RL setting. In this work, interaction affects the representations that we create (Yi we describe our initial attempt at constructing et al., 2007). Visualization interfaces should adhere to the a prototype of these ideas, through identifying human action cycle (Norman, 2013), which provides us possible features that such a system should encapsulate.