Playing Hanabi Near-Optimally

By Bruno Bouzy. 2017

30 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Economics and Econometrics, Sociology and Political Science

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

In Economics and Econometrics

In Sociology and Political Science

Citations in the corpus

The Hanabi Challenge: A New Frontier for AI Research (2019, background): “Finally, Bouzy [46] examined several heuristic players and found that combining search with the hat strategy yielded the best-performing agents.”
Bayesian Action Decoder for Deep Multi-Agent Reinforcement Learning (2018, background): “8 points in the two-player setting (Bouzy, 2017).”
Diverse Agents for Ad-Hoc Cooperation in Hanabi (2019, background): “On the other hand, a pool of very high-performing selfplay agents with sophisticated conventions (such as the ones developed in [20], [21] or learned by Reinforcement Learning in [8], [18]) might resemble more a puzzle-solving challenge”
Implicit Communication of Actionable Information in Human-AI teams (2019, background): “For example, Bouzy [8] and Cox et al. [23] use ideas from hat guessing games to get near-optimal scores.”
I See What You See: Integrating Eye Tracking into Hanabi Playing Agents (2018, background): “While this approach only works in games with 5 players, Bouzy (2017) presents an improved version that also works with fewer players. Walton-Rivers et al. (2017) present a comparison of several different approaches, including several based on Monte Carlo Tree Search (Browne et al.”
Evolving Agents for the Hanabi 2018 CIG Competition (2018, background): “To our knowledge, the only published non-hat agent that exceeds our score is the combination of Tree Search with a rule-based agent as evaluator, seen in [14], with a score of 20.”
Re-determinizing Information Set Monte Carlo Tree Search in Hanabi (2019, methodology): “one possibility would be to calculate an expected roll-forward value by enumerating all possible states, or for a random sample of them, as in Expectimax search in [23].”