… We also address the problem of automatic parameter-tuning of reinforcement learn- ing/planning
algorithms and introduce a complexity regularization-based model selection … I will not try to …

… (RL) algorithms making use of KullbackLeibler (KL) regularization as a core component …
about why KL regularization helps, so far. We study KL regularization within an approximate …

… We focus on behavior regularized approaches applied to … We introduce behavior regularized
actor critic (BRAC), a general … variants of the behavior regularized approach can yield good …

… between robustness and regularization in classification and … why regularization generalizes
well in statistical learning. … , existing regularization methods in reinforcement learning (…

… objectives, eg weight regularization, noise injection [28], or … observations, as well as
regularizing the Q-function learned by … two simple mechanisms for regularizing the value function …

… 3.2 Mixture regularization Inspired by the success of mixup in supervised learning [32], we
introduce mixture regularization (mixreg) to increase the diversity of limited training data and …

… Despite regularization being largely underutilized in deep reinforcement learning, we show
… deep neural networks to improve robustness and adaptability, we apply regularization …

… regularized using the same tools as in supervised learning (eg, weight decay, dropout),
we will use “regularization… Such regularization methods include policy regularization (penalizing …

… In this work, we propose a policy gradient algorithm, CORERL (COntrol REgularized
Reinforcement Learning), that utilizes a functional regularizer around a, typically suboptimal, control …

We propose a simple data augmentation technique that can be applied to standard model-free
reinforcement learning algorithms, enabling robust learning directly from pixels without …