Coevolution of a Backgammon Player

By Jordan B. Pollack, Alan D. Blair. 1996

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

Emergent Tool Use From Multi-Agent Autocurricula (2019, background): “Early work explored self-play using genetic algorithms [16, 17, 18, 19].”
KnightCap: A chess program that learns by combining TD(lambda) with game-tree search (1999, background): “Many authors have discussed the peculiarities of backgammon that make it particularly suitable for Temporal Difference learning with self-play [8, 6, 4].”
Abandoning Objectives: Evolution Through the Search for Novelty Alone (2011, background): “There have been several impressive successes (Chellapilla and Fogel, 1999; Hillis, 1991; Sims, 1994), but a common problem is that in practice such arms races may converge to mediocre stable states (Ficici and Pollack, 1998; Pollack et al., 1996; Watson and Pollack, 2001), cycle among various behavi”
The road to everywhere : Evolution, complexity and progress in natural and artificial systems (2008, background): “While this is a direct consequence of the Red Queen hypothesis, it is not clear how this translate in terms of global, overall progress: coevolution certainly helps maintain movement, but movement to where? It is quite possible for a population of coevolving individuals, or a group of coevolving pop”
Accelerated Neural Evolution through Cooperatively Coevolved Synapses (2008, background): “Most coevolutionary problem solving systems have concentrated on competition between species (Darwen, 1996; Pollack et al., 1996; Paredis, 1994; Miller and Cliff, 1994; Rosin, 1997).”
TDLeaf(lambda): Combining Temporal Difference Learning with Game-Tree Search (1999, background): “Temporal Difference learning or TD ( λ ) , is perhaps the best known of the reinforcement learning algorithms.”
Genetic and Evolutionary Computation – GECCO 2004 (2004, methodology): “For example, in some cases good results have been obtained by having an additional contribution from the highest quality solutions found so far (so called ‘elitist’) ants (see [8] and references therein).”
Cooperative Multi-Agent Learning: The State of the Art (2005, background): “In this situation, no learners receive sufficient feedback to improve — no matter what the dominant learner does, it always wins, and no matter what the subjugated learners do, they always lose [242, 176].”
Learning to Play Chess Using Temporal Differences (2000, background): “Many authors have discussed the peculiarities of backgammon that make it particularly suitable for Temporal Difference learning with self-play (Tesauro, 1992; Schraudolph et al., 1994; Pollack, Blair, & Land, 1996).”
Robust non-linear control through neuroevolution (2003, background): “Most coevolutionary problem solving systems have concentrated on competition between species (Darwen 1996; Pollack et al. 1996; Paredis 1994; Miller and Cliff 1994; Rosin 1997).”