The Computational Intelligence of MoGo Revealed in Taiwan's Computer Go Tournaments

  title={The Computational Intelligence of MoGo Revealed in Taiwan's Computer Go Tournaments},
  author={Chang-Shing Lee and Mei-Hui Wang and Guillaume Chaslot and Jean-Baptiste Hoock and Arpad Rimmel and Olivier Teytaud and Shang-Rong Tsai and Shun-chin Hsu and Tzung-Pei Hong},
  journal={IEEE Transactions on Computational Intelligence and AI in Games},
In order to promote computer Go and stimulate further development and research in the field, the event activities, Computational Intelligence Forum and World 9times9 Computer Go Championship, were held in Taiwan. This study focuses on the invited games played in the tournament Taiwanese Go players versus the computer program MoGo held at the National University of Tainan (NUTN), Tainan, Taiwan. Several Taiwanese Go players, including one 9-Dan (9D) professional Go player and eight amateur Go… 

Computational Intelligence Meets Game of Go

The design of games was to investigate the current level of strong programs on various board sizes, but also to monitor the human brain and to check the player strength assessment capabilities of Go programs.

Computational Intelligence Meets Game of Go @ IEEE WCCI 2012 [Society Briefs]

Since 2008, National University of Tainan (NUTN) in Taiwan and other academic organizations have hosted or organized several human vs. computer Go-related events [1, 2, 3, 4, 5] in Taiwan and in IEEE

Intelligent agents for games and computer Go

A Go game record ontology and Go board ontology schemes are presented and an ontology-based fuzzy inference system is developed to provide the regional alarm level for a Go beginner or a computer Go program in order to place the stone at the much more appropriate position.

Research on Computer Weiqi Game based on Mathematical Model

On the basis of the great improvement of computer performance, the computer Go game adopts the Monte Carlo method for dynamic evaluation, and introduces the upper bound confidence circle application tree algorithm, the search efficiency and evaluation performance have been greatly improved, and the level of computer Go program has been constantly improved.

T2FML-based adaptive assessment system for computer game of Go

A type-2 fuzzy markup language (T2FML)-based system is proposed to infer the human's rank according to simulation number, komi, and board size, and the experimental results show that the proposed approach is feasible for estimating the strength of a human.

The Game of Go @ IEEE WCCI 2010 [Society Briefs]

The game of Go is still one of the last board games where the strongest humans are still able to easily win against computers in big board games, and there are still no computer Go programs that can challenge a strong professional player in 19 3 19 games without handicap.

A Middle Game Search Algorithm Applicable to Low-Cost Personal Computer for Go

A new search algorithm based on hypothesis testing and dynamic randomization for the middle stage of the game Go, which outperforms Gnugo3.6 and Gnugos3.8 in chess power while reducing average search time and average RAM cost for one move effectively on a $19\times 19$ board is proposed.

Apply Fuzzy Markup Language to Knowledge Representation for Game of Computer Go

A model of knowledge representation including game of Go record ontology and Go board ontology based on fuzzy markup language (FML) and an FML-based fuzzy system is introduced to provide the regional alarm level for a Go beginner or a computer Go program in order to place the stone at the much more appropriate position.

Human vs. Computer Go: Review and Prospect [Discussion Forum]

The development of computational intelligence (CI) and its relative strength in comparison with human intelligence for the game of Go is discussed and it is concluded that AlphaGo beating Lee Sedol is a huge achievement in artificial intelligence (AI) based largely on CI methods.

From T2 FS-Based MoGoTW System to DyNaDF for Human and Machine Co-learning on Go

This chapter extracts the human domain knowledge from Master’s 60 games for giving the desired output and combines Particle Swarm Optimization and FML to learn the knowledge base and further infer the game results of Google AlphaGo in May 2017.



Computer Go: An AI oriented survey

Computing "Elo Ratings" of Move Patterns in the Game of Go

  • Rémi Coulom
  • Computer Science
    J. Int. Comput. Games Assoc.
  • 2007
A new Bayesian technique for supervised learning of move patterns from game records, based on a generalization of Elo ratings, which outperforms most previous pattern-learning algorithms, both in terms of mean log-evidence, and prediction rate.

Adaptive play in Texas Hold'em Poker

The main idea is to design a bot that builds beliefs on his opponent's hand that results in a global play that takes into account the opponent's strategy, and which turns out to be rather unpredictable.

Evolving an expert checkers playing program without using human expertise

Control experiments between the best evolved neural network and a program that relies on material advantage indicate the superiority of the neural network both at equal levels of look ahead and CPU time.

Games, computers, and artificial intelligence

Chips Challenging Champions: Games, Computers and Artificial Intelligence

A probabilistic Approach to Solving Crossword Puzzles and Improving Heuristic Min-Max Search by Supervised Learning.

Bandit Algorithms for Tree Search

A Bandit Algorithm for Smooth Trees (BAST) is introduced which takes into account actual smoothness of the rewards for performing efficient "cuts" of sub-optimal branches with high confidence and is illustrated on a global optimization problem of a continuous function, given noisy values.

Use of variance estimation in the multi-armed bandit problem

A logarithmic bound on the expected regret of this algorithm is proved and it is argued that the bound scales favourably with the variance of the suboptimal arms.

Finite-time Analysis of the Multiarmed Bandit Problem

This work shows that the optimal logarithmic regret is also achievable uniformly over time, with simple and efficient policies, and for all reward distributions with bounded support.

Combining online and offline knowledge in UCT

This work considers three approaches for combining offline and online value functions in the UCT algorithm, and combines these algorithms in MoGo, the world's strongest 9 x 9 Go program, where each technique significantly improves MoGo's playing strength.