Deep Reinforcement Learning (RL) is successful in solving many complex Markov Decision Processes (MDPs) problems. However, agents often face unanticipated environmental changes after deployment in the real world. These changes are often spurious and unrelated to the underlying problem, such as background shifts for visual input agents. Unfortunately, deep RL policies are usually sensitive to these changes and fail to act robustly against them. This resembles the problem of domain generalization in supervised learning. In this work, we study this problem for goal-conditioned RL agents. We propose a theoretical framework in the Block MDP setting that characterizes the generalizability of goal-conditioned policies to new environments. Under this framework, we develop a practical method PA-SkewFit that enhances domain generalization. The empirical evaluation shows that our goal-conditioned RL agent can perform well in various unseen test environments, improving by 50% over baselines.

Recently, many researchers have made successful progress in building the AI systems for MOBA-game-playing with deep reinforcement learning, such as on Dota 2 and Honor of Kings. Even though these AI systems have achieved or even exceeded human-level performance, they still suffer from the lack of policy diversity. In this paper, we propose a novel Macro-Goals Guided framework, called MGG, to learn diverse policies in MOBA games. MGG abstracts strategies as macro-goals from human demonstrations and trains a Meta-Controller to predict these macro-goals. To enhance policy diversity, MGG samples macro-goals from the Meta-Controller prediction and guides the training process towards these goals. Experimental results on the typical MOBA game Honor of Kings demonstrate that MGG can execute diverse policies in different matches and lineups, and also outperform the state-of-the-art methods over 102 heroes.

One's ability to learn a generative model of the world without supervision depends on the extent to which one can construct abstract knowledge representations that generalize across experiences. To this end, capturing an accurate statistical structure from observational data provides useful inductive biases that can be transferred to novel environments. Here, we tackle the problem of learning to control dynamical systems by applying Bayesian nonparametric methods, which is applied to solve visual servoing tasks. This is accomplished by first learning a state space representation, then inferring environmental dynamics and improving the policies through imagined future trajectories. Bayesian nonparametric models provide automatic model adaptation, which not only combats underfitting and overfitting, but also allows the model's unbounded dimension to be both flexible and computationally tractable. By employing Gaussian processes to discover latent world dynamics, we mitigate common data efficiency issues observed in reinforcement learning and avoid introducing explicit model bias by describing the system's dynamics. Our algorithm jointly learns a world model and policy by optimizing a variational lower bound of a log-likelihood with respect to the expected free energy minimization objective function. Finally, we compare the performance of our model with the state-of-the-art alternatives for continuous control tasks in simulated environments.

We propose new width-based planning and learning algorithms inspired from a careful analysis of the design decisions made by previous width-based planners. The algorithms are applied over the Atari-2600 games and our best performing algorithm, Novelty guided Critical Path Learning (N-CPL), outperforms the previously introduced width-based planning and learning algorithms $\pi$-IW(1), $\pi$-IW(1)+ and $\pi$-HIW(n, 1). Furthermore, we present a taxonomy of the Atari-2600 games according to some of their defining characteristics. This analysis of the games provides further insight into the behaviour and performance of the algorithms introduced. Namely, for games with large branching factors, and games with sparse meaningful rewards, N-CPL outperforms $\pi$-IW, $\pi$-IW(1)+ and $\pi$-HIW(n, 1).

The ability to adapt to changes in environmental contingencies is an important challenge in reinforcement learning. Indeed, transferring previously acquired knowledge to environments with unseen structural properties can greatly enhance the flexibility and efficiency by which novel optimal policies may be constructed. In this work, we study the problem of transfer learning under changes in the environment dynamics. In this study, we apply causal reasoning in the offline reinforcement learning setting to transfer a learned policy to new environments. Specifically, we use the Decision Transformer (DT) architecture to distill a new policy on the new environment. The DT is trained on data collected by performing policy rollouts on factual and counterfactual simulations from the source environment. We show that this mechanism can bootstrap a successful policy on the target environment while retaining most of the reward.

Covid-19 has shined a spotlight on many of the world's networks, from the internet to international air travel. But the supply chains that crisscross the world--the ships and trucks and trains that link factories to ports and warehouses, bringing almost everything we buy many thousands of miles from where it's produced to where it's consumed--are facing more scrutiny than they ever have. "It's fair to say that whatever you're selling, you've got a problem right now," says Jason Boyce, founder and CEO of Avenue7Media, a consulting firm that advises top Amazon sellers. Boyce says he has clients who would be turning over tens of millions of dollars a year if they could stay in stock. "We're having talks with clients every day where they're just crying," he says.

Learning in a multi-target environment without prior knowledge about the targets requires a large amount of samples and makes generalization difficult. To solve this problem, it is important to be able to discriminate targets through semantic understanding. In this paper, we propose goal-aware cross-entropy (GACE) loss, that can be utilized in a self-supervised way using auto-labeled goal states alongside reinforcement learning. Based on the loss, we then devise goal-discriminative attention networks (GDAN) which utilize the goal-relevant information to focus on the given instruction. We evaluate the proposed methods on visual navigation and robot arm manipulation tasks with multi-target environments and show that GDAN outperforms the state-of-the-art methods in terms of task success ratio, sample efficiency, and generalization. Additionally, qualitative analyses demonstrate that our proposed method can help the agent become aware of and focus on the given instruction clearly, promoting goal-directed behavior.

Future Internet involves several emerging technologies such as 5G and beyond 5G networks, vehicular networks, unmanned aerial vehicle (UAV) networks, and Internet of Things (IoTs). Moreover, future Internet becomes heterogeneous and decentralized with a large number of involved network entities. Each entity may need to make its local decision to improve the network performance under dynamic and uncertain network environments. Standard learning algorithms such as single-agent Reinforcement Learning (RL) or Deep Reinforcement Learning (DRL) have been recently used to enable each network entity as an agent to learn an optimal decision-making policy adaptively through interacting with the unknown environments. However, such an algorithm fails to model the cooperations or competitions among network entities, and simply treats other entities as a part of the environment that may result in the non-stationarity issue. Multi-agent Reinforcement Learning (MARL) allows each network entity to learn its optimal policy by observing not only the environments, but also other entities' policies. As a result, MARL can significantly improve the learning efficiency of the network entities, and it has been recently used to solve various issues in the emerging networks. In this paper, we thus review the applications of MARL in the emerging networks. In particular, we provide a tutorial of MARL and a comprehensive survey of applications of MARL in next generation Internet. In particular, we first introduce single-agent RL and MARL. Then, we review a number of applications of MARL to solve emerging issues in future Internet. The issues consist of network access, transmit power control, computation offloading, content caching, packet routing, trajectory design for UAV-aided networks, and network security issues.

Behavioral cloning has proven to be effective for learning sequential decision-making policies from expert demonstrations. However, behavioral cloning often suffers from the causal confusion problem where a policy relies on the noticeable effect of expert actions due to the strong correlation but not the cause we desire. This paper presents Object-aware REgularizatiOn (OREO), a simple technique that regularizes an imitation policy in an object-aware manner. Our main idea is to encourage a policy to uniformly attend to all semantic objects, in order to prevent the policy from exploiting nuisance variables strongly correlated with expert actions. To this end, we introduce a two-stage approach: (a) we extract semantic objects from images by utilizing discrete codes from a vector-quantized variational autoencoder, and (b) we randomly drop the units that share the same discrete code together, i.e., masking out semantic objects. Our experiments demonstrate that OREO significantly improves the performance of behavioral cloning, outperforming various other regularization and causality-based methods on a variety of Atari environments and a self-driving CARLA environment. We also show that our method even outperforms inverse reinforcement learning methods trained with a considerable amount of environment interaction.

Learned representations in deep reinforcement learning (DRL) have to extract task-relevant information from complex observations, balancing between robustness to distraction and informativeness to the policy. Such stable and rich representations, often learned via modern function approximation techniques, can enable practical application of the policy improvement theorem, even in high-dimensional continuous state-action spaces. Bisimulation metrics offer one solution to this representation learning problem, by collapsing functionally similar states together in representation space, which promotes invariance to noise and distractors. In this work, we generalize value function approximation bounds for on-policy bisimulation metrics to non-optimal policies and approximate environment dynamics. Our theoretical results help us identify embedding pathologies that may occur in practical use. In particular, we find that these issues stem from an underconstrained dynamics model and an unstable dependence of the embedding norm on the reward signal in environments with sparse rewards. Further, we propose a set of practical remedies: (i) a norm constraint on the representation space, and (ii) an extension of prior approaches with intrinsic rewards and latent space regularization. Finally, we provide evidence that the resulting method is not only more robust to sparse reward functions, but also able to solve challenging continuous control tasks with observational distractions, where prior methods fail.