Quantifying uncertainty about a policy's long-term performance is important to solve sequential decision-making tasks. We study the problem from a model-based Bayesian reinforcement learning perspective, where the goal is to learn the posterior distribution over value functions induced by parameter (epistemic) uncertainty of the Markov decision process. Previous work restricts the analysis to a few moments of the distribution over values or imposes a particular distribution shape, e.g., Gaussians. Inspired by distributional reinforcement learning, we introduce a Bellman operator whose fixed-point is the value distribution function. Based on our theory, we propose Epistemic Quantile-Regression (EQR), a model-based algorithm that learns a value distribution function that can be used for policy optimization. Evaluation across several continuous-control tasks shows performance benefits with respect to established model-based and model-free algorithms.

Ramp metering is the act of controlling on-going vehicles to the highway mainlines. Decades of practices of ramp metering have proved that ramp metering can decrease total travel time, mitigate shockwaves, decrease rear-end collisions by smoothing the traffic interweaving process, etc. Besides traditional control algorithm like ALINEA, Deep Reinforcement Learning (DRL) algorithms have been introduced to build a finer control. However, two remaining challenges still hinder DRL from being implemented in the real world: (1) some assumptions of algorithms are hard to be matched in the real world; (2) the rich input states may make the model vulnerable to attacks and data noises. To investigate these issues, we propose a Deep Q-Learning algorithm using only loop detectors information as inputs in this study. Then, a set of False Data Injection attacks and random noise attack are designed to investigate the robustness of the model. The major benefit of the model is that it can be applied to almost any ramp metering sites regardless of the road geometries and layouts. Besides outcompeting the ALINEA method, the Deep Q-Learning method also shows a good robustness through training among very different demands and geometries. For example, during the testing case in I-24 near Murfreesboro, TN, the model shows its robustness as it still outperforms ALINEA algorithm under Fast Gradient Sign Method attacks. Unlike many previous studies, the model is trained and tested in completely different environments to show the capabilities of the model.

We consider the problem of interactive decision making, encompassing structured bandits and reinforcement learning with general function approximation. Recently, Foster et al. (2021) introduced the Decision-Estimation Coefficient, a measure of statistical complexity that lower bounds the optimal regret for interactive decision making, as well as a meta-algorithm, Estimation-to-Decisions, which achieves upper bounds in terms of the same quantity. Estimation-to-Decisions is a reduction, which lifts algorithms for (supervised) online estimation into algorithms for decision making. In this paper, we show that by combining Estimation-to-Decisions with a specialized form of optimistic estimation introduced by Zhang (2022), it is possible to obtain guarantees that improve upon those of Foster et al. (2021) by accommodating more lenient notions of estimation error. We use this approach to derive regret bounds for model-free reinforcement learning with value function approximation, and give structural results showing when it can and cannot help more generally.

Offline reinforcement learning provides a viable approach to obtain advanced control strategies for dynamical systems, in particular when direct interaction with the environment is not available. In this paper, we introduce a conceptual extension for model-based policy search methods, called variable objective policy (VOP). With this approach, policies are trained to generalize efficiently over a variety of objectives, which parameterize the reward function. We demonstrate that by altering the objectives passed as input to the policy, users gain the freedom to adjust its behavior or re-balance optimization targets at runtime, without need for collecting additional observation batches or re-training.

Traffic management systems play a vital role in ensuring safe and efficient transportation on roads. However, the use of advanced technologies in traffic management systems has introduced new safety challenges. Therefore, it is important to ensure the safety of these systems to prevent accidents and minimize their impact on road users. In this survey, we provide a comprehensive review of the literature on safety in traffic management systems. Specifically, we discuss the different safety issues that arise in traffic management systems, the current state of research on safety in these systems, and the techniques and methods proposed to ensure the safety of these systems. We also identify the limitations of the existing research and suggest future research directions.

We study challenges using reinforcement learning in controlling energy systems, where apart from performance requirements, one has additional safety requirements such as avoiding blackouts. We detail how these safety requirements in real-time temporal logic can be strengthened via discretization into linear temporal logic (LTL), such that the satisfaction of the LTL formulae implies the satisfaction of the original safety requirements. The discretization enables advanced engineering methods such as synthesizing shields for safe reinforcement learning as well as formal verification, where for statistical model checking, the probabilistic guarantee acquired by LTL model checking forms a lower bound for the satisfaction of the original real-time safety requirements.

In this study, we propose a novel approach for investigating optimization performance by flexible robot coordination in automated warehouses with multi-agent reinforcement learning (MARL)-based control. Automated systems using robots are expected to achieve efficient operations compared with manual systems in terms of overall optimization performance. However, the impact of overall optimization on performance remains unclear in most automated systems due to a lack of suitable control methods. Thus, we proposed a centralized training-and-decentralized execution MARL framework as a practical overall optimization control method. In the proposed framework, we also proposed a single shared critic, trained with global states and rewards, applicable to a case in which heterogeneous agents make decisions asynchronously. Our proposed MARL framework was applied to the task selection of material handling equipment through automated order picking simulation, and its performance was evaluated to determine how far overall optimization outperforms partial optimization by comparing it with other MARL frameworks and rule-based control methods.

Function-as-a-Service (FaaS) introduces a lightweight, function-based cloud execution model that finds its relevance in applications like IoT-edge data processing and anomaly detection. While CSP offer a near-infinite function elasticity, these applications often experience fluctuating workloads and stricter performance constraints. A typical CSP strategy is to empirically determine and adjust desired function instances, "autoscaling", based on monitoring-based thresholds such as CPU or memory, to cope with demand and performance. However, threshold configuration either requires expert knowledge, historical data or a complete view of environment, making autoscaling a performance bottleneck lacking an adaptable solution.RL algorithms are proven to be beneficial in analysing complex cloud environments and result in an adaptable policy that maximizes the expected objectives. Most realistic cloud environments usually involve operational interference and have limited visibility, making them partially observable. A general solution to tackle observability in highly dynamic settings is to integrate Recurrent units with model-free RL algorithms and model a decision process as a POMDP. Therefore, in this paper, we investigate a model-free Recurrent RL agent for function autoscaling and compare it against the model-free Proximal Policy Optimisation (PPO) algorithm. We explore the integration of a LSTM network with the state-of-the-art PPO algorithm to find that under our experimental and evaluation settings, recurrent policies were able to capture the environment parameters and show promising results for function autoscaling. We further compare a PPO-based autoscaling agent with commercially used threshold-based function autoscaling and posit that a LSTM-based autoscaling agent is able to improve throughput by 18%, function execution by 13% and account for 8.4% more function instances.

This paper focuses on the impact of leveraging autonomous offensive approaches in Deep Reinforcement Learning (DRL) to train more robust agents by exploring the impact of applying adversarial learning to DRL for autonomous security in Software Defined Networks (SDN). Two algorithms, Double Deep Q-Networks (DDQN) and Neural Episodic Control to Deep Q-Network (NEC2DQN or N2D), are compared. NEC2DQN was proposed in 2018 and is a new member of the deep q-network (DQN) family of algorithms. The attacker has full observability of the environment and access to a causative attack that uses state manipulation in an attempt to poison the learning process. The implementation of the attack is done under a white-box setting, in which the attacker has access to the defender's model and experiences. Two games are played; in the first game, DDQN is a defender and N2D is an attacker, and in second game, the roles are reversed. The games are played twice; first, without an active causative attack and secondly, with an active causative attack. For execution, three sets of game results are recorded in which a single set consists of 10 game runs. The before and after results are then compared in order to see if there was actually an improvement or degradation. The results show that with minute parameter changes made to the algorithms, there was growth in the attacker's role, since it is able to win games. Implementation of the adversarial learning by the introduction of the causative attack showed the algorithms are still able to defend the network according to their strengths.

Imitation learning suffers from causal confusion. This phenomenon occurs when learned policies attend to features that do not causally influence the expert actions but are instead spuriously correlated. Causally confused agents produce low open-loop supervised loss but poor closed-loop performance upon deployment. We consider the problem of masking observed confounders in a disentangled representation of the observation space. Our novel masking algorithm leverages the usual ability to intervene in the initial system state, avoiding any requirement involving expert querying, expert reward functions, or causal graph specification. Under certain assumptions, we theoretically prove that this algorithm is conservative in the sense that it does not incorrectly mask observations that causally influence the expert; furthermore, intervening on the initial state serves to strictly reduce excess conservatism. The masking algorithm is applied to behavior cloning for two illustrative control systems: CartPole and Reacher.