Reinforcement Learning: State-of-the-Art

By Ann Nowé, Peter Vrancx, Yann-Michaël De Hauwere. 2012

546 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.

Citations in the corpus

Deep Reinforcement Learning for Autonomous Driving: A Survey (2020, background): “The goal is to ﬁnd the optimal policy π ∗ , which results in the highest expected sum of discounted rewards [19]: for all states s ∈ S , where r k = R ( s k , a k ) is the reward at time k and V π ( s ) , the ‘value function’ at state s following a policy π , is the expected ‘return’ (or ‘utility’)…”
From Bandits to Monte-Carlo Tree Search: The Optimistic Principle Applied to Optimization and Planning (2014, methodology): “[53, 116]) some parameters of the Markov decision process are initially unknown and exploration can be performed by using a Bayesian reasoning where one starts with a prior over the unknown parameters and based on the transition and reward samples observed at any time t, a posterior distribution ove”
Reinforcement learning for control: Performance, stability, and deep approximators (2018, background): “…taken, and instead overviews focus separately either on AI methods (Kaelbling et al., 1996; Sutton and Barto, 1998; Gosavi, 2009; Szepesv´ari, 2010; Bus¸oniu et al., 2010; Wiering and van Otterlo, 2012; Sut-ton and Barto, 2018) or on ADP approaches (Lewis and Vra-bie, 2009; Lewis and Liu, 2012).”
Policy Error Bounds for Model-Based Reinforcement Learning with Factored Linear Models (2016, methodology): “…to bound the policy error, while in the context of approximate linear programming, de Farias and Van Roy (2003) proposed a different technique to allow the use of weighted supremum norms, both leading to substantial further work (Busoniu et al., 2010), (Wiering and van Otterlo, 2012, Chapter 3).”
Non-parametric Policy Search with Limited Information Loss (2017, methodology): “…learning (RL) algorithms have been proposed, as reviewed by, among others, Bertsekas and Tsitsiklis (1996); Szepesvári (2010); Sutton and Barto (1998); Busoniu et al. (2010); Powell (2007); Bartlett (2003); Kaelbling et al. (1996); Wiering and Otterlo (2012); and Deisenroth et al. (2013).”
Reinforcement learning in the game of Othello: Learning against a fixed opponent and learning from self-play (2013, background): “The value of a policy π , V π ( s ) , is the expected cumulative reward that will be received when the agent follows the policy starting at state s .”
A Deep Policy Inference Q-Network for Multi-Agent Systems (2017, background): “The literature contains numerous studies of MAS modeling [7, 10, 27] and investigations of non-stationarity issues [16].”
TarMAC: Targeted Multi-Agent Communication (2018, background): “A Dec-POMDP is a multi-agent extension of a partially observable Markov decision process (Oliehoek, 2012).”
Machine learning in manufacturing: advantages, challenges, and applications (2016, background): “RL , based on sequential environmental response, emulates the process of learning of humans (Wiering & Van Otterlo, 2012).”
Planning in Discrete and Continuous Markov Decision Processes by Probabilistic Programming (2015, background): “, linear regression), nonetheless the model needs to be carefully tailored for the domain to solve [32].”
Bayesian Optimal Control of Smoothly Parameterized Systems (2015, methodology): “The effectiveness of our method is illustrated by means of a simulated example.”