Introduction to Reinforcement Learning

By Mohit Sewak. 2020

1,879 citations recorded by the source corpus.

Subfields: Management Science and Operations Research, Reinforcement Learning

How researchers see this paper

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

In Management Science and Operations Research

In Reinforcement Learning

Citations in the corpus

Constrained Policy Optimization (2017, background): “For large or continuous MDPs, solving for the exact optimal policy is intractable due to the curse of dimensionality (Sutton & Barto, 1998).”
Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning (2019, methodology): “…1 ) , ... } under a policy π , and γ ∈ [0 , 1) is the discount factor. d π ( s ) = ∞ t =0 γ t p ( s t = s | π ) represents the unnormal-ized discounted state distribution induced by the policy π (Sutton & Barto, 1998), and p ( s t = s | π ) is the likelihood of the agent being in state s after…”
Recurrent World Models Facilitate Policy Evolution (2018, background): “For many reinforcement learning (RL) problems [37, 96, 106], an artiﬁcial RL agent may also beneﬁt from a predictive model (M) of the future [95, 104] (model-based RL).”
Kernel Mean Embedding of Distributions: A Review and Beyonds (2016, methodology): “Traditional approaches for solving the Bellman equation such as dynamic programming (Bellman 2003), parametric methods (Engel et al. 2003), Monte Carlo methods (Sutton and Barto 1998), and piecewise linear and convex (PWLC)-based methods (Smallwood and Sondik 1973) have several drawbacks.”
Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems (2020, methodology): “Basic offline dynamic programming algorithms can be constructed on the basis of fitted Q-learning methods (Ernst et al., 2005), as well as policy iteration methods (Sutton and Barto, 1998).”
EDUCE: Explaining model Decisions through Unsupervised Concepts Extraction (2019, methodology): “This is a classic method in the Reinforcement Learning [25].”
Learning First-Order Symbolic Representations for Planning from the Structure of the State Space (2019, background): “In standard, tabular (model-free) reinforcement learning (RL) schemes [34], the inputs are traces made up of states, actions, and rewards, and the task is to learn a policy for maximizing (discounted) expected reward, not to learn factored symbolic representations of the actions and states.”
Deep Q-learning From Demonstrations (2017, methodology): “We adopt the standard Markov Decision Process (MDP) formalism for this work (Sutton and Barto 1998).”
Deep Reinforcement Learning and the Deadly Triad (2018, positive): “nalysing how the components of this system play a role in the emergence of the deadly triad, and in the agent’s performance. In this paper, we focus on temporal difference (TD) methods (Sutton, 1988; Sutton and Barto, 1998, 2018) to learn value functions. The value of a state, or of an action, is th”
QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulation (2018, background): “While the principles of deep reinforcement learning have been known for decades [11, 12], operationalizing them in a practical robotic learning algorithm that can generalize to new objects requires a stable and scalable algorithm and large datasets, as well as careful system design.”
Deep Reinforcement Learning for Sepsis Treatment (2017, methodology): “This work used value-iteration techniques to ﬁnd an optimal policy [13].”
Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewards (2017, methodology): “We adopt the standard Markov Decision Process (MDP) formalism for this work [15].”
Multi-agent Reinforcement Learning in Sequential Social Dilemmas (2017, background): “Modern deep reinforcement learning methods take the perspective of an agent that must learn to maximize its cumulative long-term reward through trial-and-error interactions with its environment [43, 44].”
Learning from Demonstrations for Real World Reinforcement Learning (2017, methodology): “We adopt the standard Markov Decision Process (MDP) formalism for this work Sutton & Barto (1998).”