Neurogammon: a neural-network backgammon program

By G. Tesauro. 1990

Subfields: 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 Reinforcement Learning

1. Matthew E. Taylor
2. Michael L. Littman
3. Peter Stone
4. Shimon Whiteson
5. Sergey Levine
6. Lihong Li
7. Bei Peng
8. Doina Precup

Citations in the corpus

• VAIN: Attentional Multi-agent Predictive Modeling (2017, neutral): “Approaches include [24, 25] for Chess, [26, 27, 28] for Backgammons [29] for Go.”
• Online learning control by association and reinforcement (2000, background): “The most noteworthy result is a TD-Gammon program that has learned to play Backgammon at a grandmaster level [17]–[19].”
• Stable Fitted Reinforcement Learning (1995, unknown): “G.J.”
• Residual Algorithms: Reinforcement Learning with Function Approximation (1995, methodology): “Tesauro (90,92)has shown very good results by combining TD(0) with backpropagation (and also using the more general TD(¢ )).”
• Learning behaviors via human-delivered discrete feedback: modeling implicit feedback strategies to speed up learning (2015, background): “For example, policies in continuous spaces can be represented as multilayer perceptrons, such that learning the target policy involves finding weight and bias parameters of the network that minimize or maximize some objective function based on examples of that policy [26].”