Boosting a Bridge Artificial Intelligence

@article{Ventos2017BoostingAB,
  title={Boosting a Bridge Artificial Intelligence},
  author={V{\'e}ronique Ventos and Yves Costel and Olivier Teytaud and Solene Thepaut Ventos},
  journal={2017 IEEE 29th International Conference on Tools with Artificial Intelligence (ICTAI)},
  year={2017},
  pages={1280-1287}
}
Bridge is an incomplete information game which is complex both for humans and for Computer-Bridge programs. The purpose of this paper is to present our work related to the adaptation to Bridge of a recent methodology used for boosting game Artificial Intelligence (AI) by seeking a random seed, or a probability distribution on random seeds, better than the others on a particular game. The Bridge AI Wbridge5 developed by Yves Costel has been boosted with the best seed found on the outcome of… 

Figures from this paper

Advances in computer bridge: techniques for a partial-information, communication-based game
TLDR
This thesis explores AI shortcomings in both the play and bidding phases of Bridge by exploring weaknesses in the cutting edge Monte Carlo techniques and investigating deep reinforcement learning as a method to learn how to bid.
The Game of Bridge: A Challenge for ILP
TLDR
A simple supervised learning problem in Bridge: given a ‘limit hand’, should a player bid or not, only considering his hand and the context of his decision is presented.
StarAI: Reducing incompleteness in the game of Bridge using PLP
TLDR
This work presents a methodology allowing us to model a part of card playing in Bridge using Probabilistic Logic Programming.
Who should bid higher, NS or WE, in a given Bridge dealƒ
The paper proposes a neural model for a direct comparison of the two so-called Double Dummy Bridge Problem (DDBP) instances, along with a practical use-case for determining which pair, NS or WE,
Human-Agent Cooperation in Bridge Bidding
We introduce a human-compatible reinforcement-learning approach to a cooperative game, making use of a third-party hand-coded human-compatible bot to generate initial training data and to perform
Seed Optimization Framework on Draughts
TLDR
A framework which can optimize a draughts program for competition with no modifying algorithm and no penalty when executing is provided and shows that self-learning methodology improves the strength of Scan against other competing programs.
The αμ Search Algorithm for the Game of Bridge
TLDR
Alpha, an anytime heuristic search algorithm for incomplete information games that assumes perfect information for the opponents, is applied to the game of Bridge.
Surprising Strategies Obtained by Stochastic Optimization in Partially Observable Games
TLDR
5 algorithms for tuning the parameters of strategies in the context of possibly randomized two players zero-sum games with incomplete information are compared and the seed method, which can be seen as an extremal version of coevolution, works even when nothing else works.
Optimizing αµ
TLDR
This paper optimize αµ for the game of Bridge, avoiding useless computations, and defines multiple optimizations involving Pareto fronts, and shows that these optimizations speed up the search.
Parallel Noisy Optimization in Front of Simulators : Optimism , Pessimism , Repetitions Population Control
TLDR
It is concluded that population-control methods, recently published in noisy optimization, perform greatly for continuous parameters but deceptive noise models exist and the method of seeds provides easy significant improvements though, usually, it does not scale up with the training computational power.
...
...

References

SHOWING 1-10 OF 25 REFERENCES
Automatic Bridge Bidding Using Deep Reinforcement Learning
TLDR
A flexible and pioneering bridge-bidding system, which can learn either with or without the aid of human domain knowledge, based on a novel deep reinforcement learning model, which extracts sophisticated features and learns to bid automatically based on raw card data.
DeepStack: Expert-Level Artificial Intelligence in No-Limit Poker
TLDR
DeepStack becomes the first computer program to beat professional poker players in heads-up no-limit Texas hold’em and dramatically reduces worst-case exploitability compared to the abstraction paradigm that has been favored for over a decade.
Fast Seed-Learning Algorithms for Games
TLDR
This work proposes faster variants of these algorithms, namely rectangular algorithms (fully parallel) and bandit algorithms (faster in a sequential setup) and checks the performance on several board games and card games.
The rectangular seeds of Domineering
TLDR
A methodology for boosting the computational intelligence of randomized game-playing programs is modified by working on rectangular, rather than square, matrices; and it is applied to the Domineering game.
Neural networks for contract bridge bidding
TLDR
It is shown that a multilayer feedforward neural network can be trained to learn to make an opening bid with a new hand, and the need for a hierarchical architecture to deal with bids at all levels is discussed.
Contract Bridge Bidding by Learning
  • C. HoHsuan-Tien Lin
  • Computer Science
    AAAI Workshop: Computer Poker and Imperfect Information
  • 2015
TLDR
A novel learning framework to let a computer program learn its own bidding decisions is proposed and it is found that it performs competitively to the champion computer bridge program that mimics human bidding decisions.
Learning to bid in bridge
TLDR
A new decision-making algorithm that allows models to be used for both opponent agents and partners, while utilizing a novel model-based Monte Carlo sampling method to overcome the problem of hidden information is presented.
Mastering the game of Go with deep neural networks and tree search
TLDR
Using this search algorithm, the program AlphaGo achieved a 99.8% winning rate against other Go programs, and defeated the human European Go champion by 5 games to 0.5, the first time that a computer program has defeated a human professional player in the full-sized game of Go.
Bridge Bidding with Imperfect Information
  • L. DeLoozeJ. Downey
  • Computer Science
    2007 IEEE Symposium on Computational Intelligence and Games
  • 2007
TLDR
It is shown that a special form of a neural network, called a self-organizing map (SOM), can be used to effectively bid no trump hands and is an ideal mechanism for modeling the imprecise and ambiguous nature of the game.
The State of Automated Bridge Play
TLDR
The game of Bridge provides a number of research areas to AI researchers due to the many components that constitute the game, particularly double-dummy play, but researchers have made much progress in each of these sub-fields over the years, but are yet to produce a consistent expert level player.
...
...