Reinforcement learning - an introduction, 2nd Edition

By Richard S. Sutton, Andrew Barto. 2018

2,091 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Reinforcement Learning

Citations in the corpus

Model-Based Reinforcement Learning for Atari (2019, neutral): “One of the exciting beneﬁts of model-based learning is the promise to substantially improve sample efﬁciency of deep reinforcement learning (see Chapter 8 in Sutton & Barto (2017)).”
Distributed Prioritized Experience Replay (2018, positive): “Finally, by learning off-policy (cf. Sutton & Barto, 1998; 2017), we can further take advantage of Ape-X’s ability to combine data from many distributed actors, by giving the different actors different exploration policies, broadening the diversity of the experience they jointly encounter.”
Deep Reinforcement Learning for Autonomous Driving: A Survey (2020, positive): “An advantage of this separation is that the target policy may be deterministic (greedy), while the behavior policy can continue to sample all possible actions, [1].”
D4RL: Datasets for Deep Data-Driven Reinforcement Learning (2020, neutral): “While oﬀ-policy RL algorithms can in principle be used in the oﬄine setting, the lack of ability to collect additional data can break assumptions required for convergence, such as in online Q-learning [Sutton and Barto, 2018].”
GradientDICE: Rethinking Generalized Offline Estimation of Stationary Values (2020, neutral): “A key challenge in reinforcement learning (RL, Sutton & Barto 2018) is off-policy evaluation (Precup et al., 2001; Maei, 2011; Jiang & Li, 2015; Sutton & Barto, 2018; Liu et al., 2018; Nachum et al., 2019; Zhang et al., 2020), where we want to estimate the performance of a target policy (average…”
Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction (2019, neutral): “Actor-critic methods [36, 14, 19] address this by additionally learning a policy π θ that maximizes the Q -function.”
A survey and critique of multiagent deep reinforcement learning (2018, positive): “While different techniques and algorithms were used in the above scenarios, in general, they are all a combination of techniques from two main areas: reinforcement learning (RL) [315] and deep learning [184,281].”
Deep Reinforcement Learning: An Overview (2017, positive): “See Chapter 16 in Sutton and Barto (2017) for a detailed and intuitive description of AlphaGo.”
Wireless Networks Design in the Era of Deep Learning: Model-Based, AI-Based, or Both? (2019, neutral): “For a dedicated and comprehensive treatment of the reinforcement learning framework, we refer the reader to [97].”
Eligibility Traces and Plasticity on Behavioral Time Scales: Experimental Support of NeoHebbian Three-Factor Learning Rules (2018, positive): “An excellent modern summary of Reinforcement Learning Algorithms and their historical predecessors can be found in (Sutton and Barto, 2018).”
A survey of multi-agent deep reinforcement learning with communication (2022, positive): “Moreover, the Q-function in Equation 3 can be replaced by average returns over episodes to form REINFORCE algorithms [26], or by an estimated value function to form actor-critic algorithms [35, 36].”
Motion Planning for Autonomous Driving: The State of the Art and Future Perspectives (2023, positive): “The goal is to find the optimal policy π∗, which results in the highest expected sum of discounted rewards [81]:”
A comprehensive survey on machine learning for networking: evolution, applications and research opportunities (2018, positive): “MDP has been essential for the development of prevailing RL techniques [435].”
Deep reinforcement learning: a survey (2020, positive): “Combining the advantages of model-free RL, Dyna-Q (Sutton and Barto, 2018) is a classical integrated architecture where a model is leveraged to update Q-values.”