Reinforcement Learning and Its Relationship to Supervised Learning

By Jennie Si, Andrew Barto, Warrren B Powell, Donald C. Wunsch II. 2004

106 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 Artificial Intelligence

Citations in the corpus

Towards Continual Reinforcement Learning: A Review and Perspectives (2020, background): “the RL framework as highlighted by (Barto and Dietterich, 2004).”
Training Agents using Upside-Down Reinforcement Learning (2019, background): “On the possibility of turning an RL problem completely into an SL problem, Barto and Dietterich [2004] surmised: “In general, there is no way to do this.”
Partial Policy-Based Reinforcement Learning for Anatomical Landmark Localization in 3D Medical Images (2018, unknown): “[41].”
Reinforcement learning and adaptive dynamic programming for feedback control (2009, background): “It has been shown by [Leake and Liu 1967], [Howard 1960] and others that this algorithm converges under certain conditions to the optimal value and control policy, that is, to the solution of (11), (12).”
Provably safe and robust learning-based model predictive control (2011, unknown): “el is not a good representation of the true dynamics. It is known that the performance of a learning-based controller can be arbitrarily bad if the learned model does not exactly match the true model [15]; imposing the constraints on the nominal model, instead of the learned model, protects against ”
Online Learning: A Comprehensive Survey (2018, background): “Reinforcement learning is different from supervised learning [310] in that the goal of supervised learning is to reconstruct an unknown function f that can assign the desired output values y to input data x; while the goal of RL is to find the input (policy/action) x that gives the maximum reward R(”