Aspects of the Cooperative Card Game Hanabi

By Mark J. H. van den Bergh, A. Hommelberg, W. Kosters, Flora M. Spieksma. 2016

29 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.

The Hanabi Challenge: A New Frontier for AI Research (2019, methodology): “described fixed rules whose trigger thresholds were optimized by manual tuning via simulation [46]; the action selection is also improved during play using Monte Carlo planning.”
Playing Hanabi Near-Optimally (2017, background): “Van den Berghe [8] describes experiments with three players and five cards per player.”
Evolving Agents for the Hanabi 2018 CIG Competition (2018, background): “[9] optimize a rule-based agent by exhaustively searching 48 possible agents obtained by selecting one of 4 possible hint heuristics, 4 discard heuristics and 3 thresholds for playability of a card.”
Re-determinizing Information Set Monte Carlo Tree Search in Hanabi (2019, methodology): “Our solution to this problem, which we term Redeterminizing IS-MCTS (RIS-MCTS) is to re-determinize the game-state at every node in a similar fashion to the re-shuffle of [21].”
Implicit Communication of Actionable Information in Human-AI teams (2019, background): “However, many rule-based approaches apply techniques that would be entirely opaque to human teammates [12, 83].”