Reward shaping has been demonstrated to be an effective technique for accelerating the learning process of reinforcement learning (RL) agents. While successful in empirical applications, the design of a good shaping function is less well understood in principle and thus often relies on domain expertise and manual design. To overcome this limitation, we propose a novel automated approach for designing reward functions from offline data, possibly contaminated with the unobserved confounding bias.We propose to use causal state value upper bounds calculated from offline datasets as a conservative optimistic estimation of the optimal state value, which is then used as state potentials in Potential-Based Reward Shaping (PBRS). When applying our shaping function to a model-free learner based on UCB principles, we show that it enjoys a better gap-dependent regret bound than the learner without shaping. To the best of our knowledge, this is the first gap-dependent regret bound for PBRS in model-free learning with online exploration. Simulations support the theoretical findings.

A key task in Artificial Intelligence is learning effective policies for controlling agents in unknown environments to optimize performance measures. Off-policy learning methods, like Q-learning, allow learners to make optimal decisions based on past experiences. This paper studies off-policy learning from biased data in complex and high-dimensional domains where \emph{unobserved confounding} cannot be ruled out a priori. Building on the well-celebrated Deep Q-Network (DQN), we propose a novel deep reinforcement learning algorithm robust to confounding biases in observed data. Specifically, our algorithm attempts to find a safe policy for the worst-case environment compatible with the observations. We apply our method to twelve confounded Atari games, and find that it consistently dominates the standard DQN in all games where the observed input to the behavioral and target policies mismatch and unobserved confounders exist.

A pervasive challenge in Reinforcement Learning (RL) is the ``curse of dimensionality'' which is the exponential growth in the state-action space when optimizing a high-dimensional target task. The framework of curriculum learning trains the agent in a curriculum composed of a sequence of related and more manageable source tasks. The expectation is that when some optimal decision rules are shared across source tasks and the target task, the agent could more quickly pick up the necessary skills to behave optimally in the environment, thus accelerating the learning process. However, this critical assumption of invariant optimal decision rules does not necessarily hold in many practical applications, specifically when the underlying environment contains unobserved confounders. This paper studies the problem of curriculum RL through causal lenses. We derive a sufficient graphical condition characterizing causally aligned source tasks, i.e., the invariance of optimal decision rules holds. We further develop an efficient algorithm to generate a causally aligned curriculum, provided with qualitative causal knowledge of the target environment. Finally, we validate our proposed methodology through experiments in confounded environments.

In this work, we propose and explore Deep Graph Value Network (DeepGV) as a promising method to work around sample complexity in deep reinforcement-learning agents using a message-passing mechanism. The main idea is that the agent should be guided by structured non-neural-network algorithms like dynamic programming. According to recent advances in algorithmic alignment, neural networks with structured computation procedures can be trained efficiently. We demonstrate the potential of graph neural network in supporting sample efficient learning by showing that Deep Graph Value Network can outperform unstructured baselines by a large margin in solving Markov Decision Process (MDP). We believe this would open up a new avenue for structured agents design. See https://github.com/drmeerkat/Deep-Graph-Value-Network for the code.

While having achieved great success in rich real-life applications, deep neural network (DNN) models have long been criticized for their vulnerability to adversarial attacks. Tremendous research efforts have been dedicated to mitigating the threats of adversarial attacks, but the essential trait of adversarial examples is not yet clear, and most existing methods are yet vulnerable to hybrid attacks and suffer from counterattacks. In light of this, in this paper, we first reveal a gradient-based correlation between sensitivity analysis-based DNN interpreters and the generation process of adversarial examples, which indicates the Achilles's heel of adversarial attacks and sheds light on linking together the two long-standing challenges of DNN: fragility and unexplainability. We then propose an interpreter-based ensemble framework called X-Ensemble for robust adversary defense. X-Ensemble adopts a novel detection-rectification process and features in building multiple sub-detectors and a rectifier upon various types of interpretation information toward target classifiers. Moreover, X-Ensemble employs the Random Forests (RF) model to combine sub-detectors into an ensemble detector for adversarial hybrid attacks defense. The non-differentiable property of RF further makes it a precious choice against the counterattack of adversaries. Extensive experiments under various types of state-of-the-art attacks and diverse attack scenarios demonstrate the advantages of X-Ensemble to competitive baseline methods.