We begin with the hypothesis that a model-free agent whose representations are predictive of properties of future states (beyond expected rewards) will be more capable of solving and adapting to new RL problems. To test that hypothesis, we introduce an objective based on Deep InfoMax (DIM) which trains the agent to predict the future by maximizing the mutual information between its internal representation of successive timesteps. We test our approach in several synthetic settings, where it successfully learns representations that are predictive of the future. Finally, we augment C51, a strong RL baseline, with our temporal DIM objective and demonstrate improved performance on a continual learning task and on the recently introduced Procgen environment.

In many reinforcement learning (RL) problems, it takes some time until a taken action by the agent reaches its maximum effect on the environment and consequently the agent receives the reward corresponding to that action by a delay called action-effect delay. Such delays reduce the performance of the learning algorithm and increase the computational costs, as the reinforcement learning agent values the immediate rewards more than the future reward that is more related to the taken action. This paper addresses this issue by introducing an applicable enhanced Q-learning method in which at the beginning of the learning phase, the agent takes a single action and builds a function that reflects the environments response to that action, called the reflexive $\gamma$ - function. During the training phase, the agent utilizes the created reflexive $\gamma$- function to update the Q-values. We have applied the developed method to a structural control problem in which the goal of the agent is to reduce the vibrations of a building subjected to earthquake excitations with a specified delay. Seismic control problems are considered as a complex task in structural engineering because of the stochastic and unpredictable nature of earthquakes and the complex behavior of the structure. Three scenarios are presented to study the effects of zero, medium, and long action-effect delays and the performance of the Enhanced method is compared to the standard Q-learning method. Both RL methods use neural network to learn to estimate the state-action value function that is used to control the structure. The results show that the enhanced method significantly outperforms the performance of the original method in all cases, and also improves the stability of the algorithm in dealing with action-effect delays.

Curriculum learning (CL) is a training strategy that trains a machine learning model from easier data to harder data, which imitates the meaningful learning order in human curricula. As an easy-to-use plug-in tool, the CL strategy has demonstrated its power in improving the generalization capacity and convergence rate of various models in a wide range of scenarios such as computer vision and natural language processing, etc. In this survey article, we comprehensively review CL from various aspects including motivations, definitions, theories, and applications. We discuss works on curriculum learning within a general CL framework, elaborating on how to design a manually predefined curriculum or an automatic curriculum. In particular, we summarize existing CL designs based on the general framework of Difficulty Measurer + Training Scheduler and further categorize the methodologies for automatic CL into four groups, i.e., Self-paced Learning, Transfer Teacher, RL Teacher, and Other Automatic CL. Finally, we present brief discussions on the relationships between CL and other methods, and point out potential future research directions deserving further investigations.

Value Iteration Networks (VINs) have emerged as a popular method to incorporate planning algorithms within deep reinforcement learning, enabling performance improvements on tasks requiring long-range reasoning and understanding of environment dynamics. This came with several limitations, however: the model is not incentivised in any way to perform meaningful planning computations, the underlying state space is assumed to be discrete, and the Markov decision process (MDP) is assumed fixed and known. We propose eXecuted Latent Value Iteration Networks (XLVINs), which combine recent developments across contrastive self-supervised learning, graph representation learning and neural algorithmic reasoning to alleviate all of the above limitations, successfully deploying VIN-style models on generic environments. XLVINs match the performance of VIN-like models when the underlying MDP is discrete, fixed and known, and provides significant improvements to model-free baselines across three general MDP setups.

Maximum target coverage by adjusting the orientation of distributed sensors is an important problem in directional sensor networks (DSNs). This problem is challenging as the targets usually move randomly but the coverage range of sensors is limited in angle and distance. Thus, it is required to coordinate sensors to get ideal target coverage with low power consumption, e.g. no missing targets or reducing redundant coverage. To realize this, we propose a Hierarchical Target-oriented Multi-Agent Coordination (HiT-MAC), which decomposes the target coverage problem into two-level tasks: targets assignment by a coordinator and tracking assigned targets by executors. Specifically, the coordinator periodically monitors the environment globally and allocates targets to each executor. In turn, the executor only needs to track its assigned targets. To effectively learn the HiT-MAC by reinforcement learning, we further introduce a bunch of practical methods, including a self-attention module, marginal contribution approximation for the coordinator, goal-conditional observation filter for the executor, etc. Empirical results demonstrate the advantage of HiT-MAC in coverage rate, learning efficiency,and scalability, comparing to baselines. We also conduct an ablative analysis on the effectiveness of the introduced components in the framework.

This paper presents an inverse reinforcement learning (IRL) framework for Bayesian stopping time problems. By observing the actions of a Bayesian decision maker, we provide a necessary and sufficient condition to identify if these actions are consistent with optimizing a cost function; then we construct set valued estimates of the cost function. To achieve this IRL objective, we use novel ideas from Bayesian revealed preferences stemming from microeconomics. To illustrate our IRL scheme,we consider two important examples of stopping time problems, namely, sequential hypothesis testing and Bayesian search. Finally, for finite datasets, we propose an IRL detection algorithm and give finite sample bounds on its error probabilities. Also we discuss how to identify $\epsilon$-optimal Bayesian decision makers and perform IRL.

Model-based reinforcement learning is a framework in which an agent learns an environment model, makes planning and decision-making in this model, and finally interacts with the real environment. Model-based reinforcement learning has high sample efficiency compared with model-free reinforcement learning, and shows great potential in the real-world application. However, model-based reinforcement learning has been plagued by dynamics bottleneck. Dynamics bottleneck is the phenomenon that when the timestep to interact with the environment increases, the reward of the agent falls into the local optimum instead of increasing. In this paper, we analyze and explain how the coupling relationship between model and policy causes the dynamics bottleneck and shows improving the exploration ability of the agent can alleviate this issue. We then propose a new planning algorithm called Maximum Entropy Cross-Entropy Method (MECEM). MECEM can improve the agent's exploration ability by maximizing the distribution of action entropy in the planning process. We conduct experiments on fourteen well-recognized benchmark environments such as HalfCheetah, Ant and Swimmer. The results verify that our approach obtains the state-of-the-art performance on eleven benchmark environments and can effectively alleviate dynamics bottleneck on HalfCheetah, Ant and Walker2D.

We study episodic reinforcement learning in non-stationary linear (a.k.a. low-rank) Markov Decision Processes (MDPs), i.e, both the reward and transition kernel are linear with respect to a given feature map and are allowed to evolve either slowly or abruptly over time. For this problem setting, we propose OPT-WLSVI an optimistic model-free algorithm based on weighted least squares value iteration which uses exponential weights to smoothly forget data that are far in the past. We show that our algorithm, when competing against the best policy at each time, achieves a regret that is upped bounded by $\widetilde{\mathcal{O}}(d^{7/6}H^2 \Delta^{1/3} K^{2/3})$ where $d$ is the dimension of the feature space, $H$ is the planning horizon, $K$ is the number of episodes and $\Delta$ is a suitable measure of non-stationarity of the MDP. This is the first regret bound for non-stationary reinforcement learning with linear function approximation.

Goal-conditioned hierarchical reinforcement learning (HRL) is a promising approach for scaling up reinforcement learning (RL) techniques. However, it often suffers from training inefficiency as the action space of the high-level, i.e., the goal space, is often large. Searching in a large goal space poses difficulties for both high-level subgoal generation and low-level policy learning. In this paper, we show that this problem can be effectively alleviated by restricting the high-level action space from the whole goal space to a $k$-step adjacent region of the current state using an adjacency constraint. We theoretically prove that the proposed adjacency constraint preserves the optimal hierarchical policy in deterministic MDPs, and show that this constraint can be practically implemented by training an adjacency network that can discriminate between adjacent and non-adjacent subgoals. Experimental results on discrete and continuous control tasks show that incorporating the adjacency constraint improves the performance of state-of-the-art HRL approaches in both deterministic and stochastic environments.

Symbolic knowledge (e.g., entities, relations, and facts in a knowledge graph) has become an increasingly popular component of neural-symbolic models applied to machine learning tasks, such as question answering and recommender systems. Besides improving downstream performance, these symbolic structures (and their associated attention weights) are often used to help explain the model's predictions and provide "insights" to practitioners. In this paper, we question the faithfulness of such symbolic explanations. We demonstrate that, through a learned strategy (or even simple heuristics), one can produce deceptively perturbed symbolic structures which maintain the downstream performance of the original structure while significantly deviating from the original semantics. In particular, we train a reinforcement learning policy to manipulate relation types or edge connections in a knowledge graph, such that the resulting downstream performance is maximally preserved. Across multiple models and tasks, our approach drastically alters knowledge graphs with little to no drop in performance. These results raise doubts about the faithfulness of explanations provided by learned symbolic structures and the reliability of current neural-symbolic models in leveraging symbolic knowledge.