Learning-based approaches for solving large sequential decision making problems have become popular in recent years. The resulting agents perform differently and their characteristics depend on those of the underlying learning approach. Here, we consider a benchmark planning problem from the reinforcement learning domain, the Racetrack, to investigate the properties of agents derived from different deep (reinforcement) learning approaches. We compare the performance of deep supervised learning, in particular imitation learning, to reinforcement learning for the Racetrack model. We find that imitation learning yields agents that follow more risky paths. In contrast, the decisions of deep reinforcement learning are more foresighted, i.e., avoid states in which fatal decisions are more likely. Our evaluations show that for this sequential decision making problem, deep reinforcement learning performs best in many aspects even though for imitation learning optimal decisions are considered.

In this paper, several techniques for learning game state evaluation functions by reinforcement are proposed. The first is a generalization of tree bootstrapping (tree learning): it is adapted to the context of reinforcement learning without knowledge based on non-linear functions. With this technique, no information is lost during the reinforcement learning process. The second is a modification of minimax with unbounded depth extending the best sequences of actions to the terminal states. This modified search is intended to be used during the learning process. The third is to replace the classic gain of a game (+1 / -1) with a reinforcement heuristic. We study particular reinforcement heuristics such as: quick wins and slow defeats ; scoring ; mobility or presence. The four is another variant of unbounded minimax, which plays the safest action instead of playing the best action. This modified search is intended to be used after the learning process. The five is a new action selection distribution. The conducted experiments suggest that these techniques improve the level of play. Finally, we apply these different techniques to design program-players to the game of Hex (size 11 and 13) surpassing the level of Mohex 2.0 with reinforcement learning from self-play without knowledge. At Hex size 11 (without swap), the program-player reaches the level of Mohex 3HNN.

We explore value-based multi-agent reinforcement learning (MARL) in the popular paradigm of centralized training with decentralized execution (CTDE). CTDE requires the consistency of the optimal joint action selection with optimal individual action selections, which is called the IGM (Individual-Global-Max) principle. However, in order to achieve scalability, existing MARL methods either limit representation expressiveness of their value function classes or relax the IGM consistency, which may lead to poor policies or even divergence. This paper presents a novel MARL approach, called duPLEX dueling multi-agent Q-learning (QPLEX), that takes a duplex dueling network architecture to factorize the joint value function. This duplex dueling architecture transforms the IGM principle to easily realized constraints on advantage functions and thus enables efficient value function learning. Theoretical analysis shows that QPLEX solves a rich class of tasks. Empirical experiments on StarCraft II unit micromanagement tasks demonstrate that QPLEX significantly outperforms state-of-the-art baselines in both online and offline task settings, and also reveal that QPLEX achieves high sample efficiency and can benefit from offline datasets without additional exploration.

We present a review and taxonomy of 200 models from the literature on driver behavior modeling. We begin by introducing a mathematical formulation based on the partially observable stochastic game, which serves as a common framework for comparing and contrasting different driver models. Our taxonomy is constructed around the core modeling tasks of state estimation, intention estimation, trait estimation, and motion prediction, and also discusses the auxiliary tasks of risk estimation, anomaly detection, behavior imitation and microscopic traffic simulation. Existing driver models are categorized based on the specific tasks they address and key attributes of their approach.

Learning from demonstration is widely used as an efficient way for robots to acquire new skills. However, it typically requires that demonstrations provide full access to the state and action sequences. In contrast, learning from observation offers a way to utilize unlabeled demonstrations (e.g., video) to perform imitation learning. One approach to this is behavioral cloning from observation (BCO). The original implementation of BCO proceeds by first learning an inverse dynamics model and then using that model to estimate action labels, thereby reducing the problem to behavioral cloning. However, existing approaches to BCO require a large number of initial interactions in the first step. Here, we provide a novel theoretical analysis of BCO, introduce a modification BCO*, and show that in the semi-supervised setting, BCO* can concurrently improve both its estimate for the inverse dynamics model and the expert policy. This result allows us to eliminate the dependence on initial interactions and dramatically improve the sample complexity of BCO. We evaluate the effectiveness of our algorithm through experiments on various benchmark domains. The results demonstrate that concurrent training not only improves over the performance of BCO but also results in performance that is competitive with state-of-the-art imitation learning methods such as GAIL and Value-Dice.

This paper introduces two simple techniques to improve off-policy Reinforcement Learning (RL) algorithms. First, we formulate off-policy RL as a stochastic proximal point iteration. The target network plays the role of the variable of optimization and the value network computes the proximal operator. Second, we exploits the two value functions commonly employed in state-of-the-art off-policy algorithms to provide an improved action value estimate through bootstrapping with limited increase of computational resources. Further, we demonstrate significant performance improvement over state-of-the-art algorithms on standard continuous-control RL benchmarks.

With the constant demand for connectivity at an all-time high, Network Service Providers (NSPs) are required to optimize their networks to cope with rising capital and operational expenditures required to meet the growing connectivity demand. A solution to this challenge was presented through Network Function Virtualization (NFV). As network complexity increases and futuristic networks take shape, NSPs are required to incorporate an increasing amount of operational efficiency into their NFV-enabled networks. One such technique is Machine Learning (ML), which has been applied to various entities in NFV-enabled networks, most notably in the NFV Orchestrator. While traditional ML provides tremendous operational efficiencies, including real-time and high-volume data processing, challenges such as privacy, security, scalability, transferability, and concept drift hinder its widespread implementation. Through the adoption of Advanced Intelligence techniques such as Reinforcement Learning and Federated Learning, NSPs can leverage the benefits of traditional ML while simultaneously addressing the major challenges traditionally associated with it. This work presents the benefits of adopting these advanced techniques, provides a list of potential use cases and research topics, and proposes a bottom-up micro-functionality approach to applying these methods of Advanced Intelligence to NFV Management and Orchestration.

Deep reinforcement learning (DRL) is becoming a prevalent and powerful methodology to address the artificial intelligent problems. Owing to its tremendous potentials in self-learning and self-improvement, DRL is broadly serviced in many research fields. This article conducted a comprehensive comparison of multiple DRL approaches on the freeway decision-making problem for autonomous vehicles. These techniques include the common deep Q learning (DQL), double DQL (DDQL), dueling DQL, and prioritized replay DQL. First, the reinforcement learning (RL) framework is introduced. As an extension, the implementations of the above mentioned DRL methods are established mathematically. Then, the freeway driving scenario for the automated vehicles is constructed, wherein the decision-making problem is transferred as a control optimization problem. Finally, a series of simulation experiments are achieved to evaluate the control performance of these DRL-enabled decision-making strategies. A comparative analysis is realized to connect the autonomous driving results with the learning characteristics of these DRL techniques.

Despite the significant progress of deep reinforcement learning (RL) in solving sequential decision making problems, RL agents often overfit to training environments and struggle to adapt to new, unseen environments. This prevents robust applications of RL in real world situations, where system dynamics may deviate wildly from the training settings. In this work, our primary contribution is to propose an information theoretic regularization objective and an annealing-based optimization method to achieve better generalization ability in RL agents. We demonstrate the extreme generalization benefits of our approach in different domains ranging from maze navigation to robotic tasks; for the first time, we show that agents can generalize to test parameters more than 10 standard deviations away from the training parameter distribution. This work provides a principled way to improve generalization in RL by gradually removing information that is redundant for task-solving; it opens doors for the systematic study of generalization from training to extremely different testing settings, focusing on the established connections between information theory and machine learning.

One difficulty in using artificial agents for human-assistive applications lies in the challenge of accurately assisting with a person's goal(s). Existing methods tend to rely on inferring the human's goal, which is challenging when there are many potential goals or when the set of candidate goals is difficult to identify. We propose a new paradigm for assistance by instead increasing the human's ability to control their environment, and formalize this approach by augmenting reinforcement learning with human empowerment. This task-agnostic objective preserves the person's autonomy and ability to achieve any eventual state. We test our approach against assistance based on goal inference, highlighting scenarios where our method overcomes failure modes stemming from goal ambiguity or misspecification. As existing methods for estimating empowerment in continuous domains are computationally hard, precluding its use in real time learned assistance, we also propose an efficient empowerment-inspired proxy metric. Using this, we are able to successfully demonstrate our method in a shared autonomy user study for a challenging simulated teleoperation task with human-in-the-loop training.