SVM and Pattern-Enriched Common Fate Graphs for the Game of Go

Go is an ancient board game that poses unique opportunities and challenges for AI and machine learning. Here we develop a machine learning approach to Go, and related board games, focusing primarily on the problem of learning a good evaluation function in a scalable way. Scalability is essential at multiple levels, from the library of local tactical patterns, to the integration of patterns across the board, to the size of the board itself. The system we propose is capable of automatically learning the propensity of local patterns from a library of games. Propensity and other local tactical information are fed into a recursive neural network, derived from a Bayesian network architecture. The network integrates local information across the board and produces local outputs that represent local territory ownership probabilities. The aggregation of these probabilities provides an effective strategic evaluation function that is an estimate of the expected area at the end (or at other stages) of the game. Local area targets for training can be derived from datasets of human games. A system trained using only 9 × 9 amateur game data performs surprisingly well on a test set derived from 19 × 19 professional game data. Possible directions for further improvements are briefly discussed.

Go is an ancient board game that poses unique opportunities and challenges for artificial intelligence. Currently, there are no computer Go programs that can play at the level of a good human player. However, the emergence of large repositories of games is opening the door for new machine learning approaches to address this challenge. Here we develop a machine learning approach to Go, and related board games, focusing primarily on the problem of learning a good evaluation function in a scalable way. Scalability is essential at multiple levels, from the library of local tactical patterns, to the integration of patterns across the board, to the size of the board itself. The system we propose is capable of automatically learning the propensity of local patterns from a library of games. Propensity and other local tactical information are fed into recursive neural networks, derived from a probabilistic Bayesian network architecture. The recursive neural networks in turn integrate local information across the board in all four cardinal directions and produce local outputs that represent local territory ownership probabilities. The aggregation of these probabilities provides an effective strategic evaluation function that is an estimate of the expected area at the end, or at various other stages, of the game. Local area targets for training can be derived from datasets of games played by human players. In this approach, while requiring a learning time proportional to N 4 , skills learned on a board of size N 2 can easily be transferred to boards of other sizes. A system trained using only 9 × 9 amateur game data performs surprisingly well on a test set derived from 19 × 19 professional game data. Possible directions for further improvements are briefly discussed.

2015 IEEE Conference on Computational Intelligence and Games (CIG), 2015

We propose a way of extracting and aggregating permove evaluations from sets of Go game records. The evaluations capture different aspects of the games such as played patterns or statistic of sente/gote sequences. Using machine learning algorithms, the evaluations can be utilized to predict different relevant target variables. We apply this methodology to predict the strength and playing style of the player (e.g. territoriality or aggressivity) with good accuracy. We propose a number of possible applications including aiding in Go study, seeding realwork ranks of internet players or tuning of Go-playing programs.

… of the IJCAI Workshop on General …, 2011

General Game Playing (GGP) is concerned with the development of programs that are able to play previously unknown games well. The main problem such a player is faced with is to come up with a good heuristic evaluation function automatically. Part of these heuristics are distance measures used to estimate, e.g., the distance of a pawn towards the promotion rank. However, current distance heuristics in GGP are based on too specific detection patterns as well as expensive internal simulations, they are limited to the scope of totally ordered domains and/or they apply a uniform Manhattan distance heuristics regardless of the move pattern of the object involved. In this paper we describe a method to automatically construct distance measures by analyzing the game rules. The presented method is an improvement to all previously presented distance estimation methods, because it is not limited to specific structures, such as, game boards. Furthermore, the constructed distance measures are admissible. We demonstrate how to use the distance measures in an evaluation function of a general game player and show the effectiveness of our approach by comparing with a state-of-the-art player.

Proceedings of the 2nd Annual European …, 2001

In this paper, a three-component architecture of a learning environment for Go is sketched, which can be applied to any two-player, deterministic, full information, partizan, combinatorial game. The architecture called HUGO has natural and human-like reasoning components. Its most abstract component deals with the selection of subgames of Go. The second component is concerned with initiative. The notion of gote no sente (a move that loses initiative but creates new lines of play that will hold initiative) is formalized. In the third component, game values are computed with a new kind of ®-¯algorithm based on fuzzy, partial ordering. Our approach leaves some valuable control parameters and o¤ers ways to apply further machine learning techniques.

International Conference on Computational Intelligence for Modelling, Control and Automation and International Conference on Intelligent Agents, Web Technologies and Internet Commerce (CIMCA-IAWTIC'06)

We present an approach for improving computer Go programs called CSDTA (Canonical Sequence Directed Tactics Analyzer), which analyzes the local tactics on a quadrant of the standard Go board based on a collection of canonical sequences (Joseki). We collect 1278 canonical sequences and their deviations in our system. Instead of trivially matching the current game and the collected canonical sequences, we define a notion of similar sequences with respect to the current game. This paper also explains how to extract the most suitable move from the candidate sequences for the next move. The simplicity of our method and its positive outcome make our approach suitable to be intergraded in a complete computer Go program for foreseeable improvement.

1994

A rst order system, PAL, that can learn Chess patterns in the form of Horn clauses from simple example descriptions and general purpose knowledge about Chess is described. This is the rst time that Chess patterns which can be used for over-the-board play have been learned. To test if the patterns learned by PAL can be used to play, a simple playing strategy for the King and Rook against King (KRK) endgame was constructed with patterns learned by PAL. Limitations of PAL in particular, and rstorder systems in general, are exposed in domains like Chess where a large number of background de nitions may be required for induction. Conclusions and future research directions are given.

In ranking problems, the goal is to learn a ranking function r(x) ∈ R from labeled pairs x, x ′ of input points. In this paper, we consider the related comparison problem, where the label y ∈ {−1, 0, 1} indicates which element of the pair is better (y = −1 or 1), or if there is no significant difference (y = 0). We cast the learning problem as a margin maximization, and show that it can be solved by converting it to a standard SVM. We use simulated nonlinear patterns and a real learning to rank sushi data set to show that our proposed SVMcompare algorithm outperforms SVMrank when there are equality y = 0 pairs. In addition, we show that SVMcompare outperforms the ELO rating system when predicting the outcome of chess matches.

Eighth International Conference on Hybrid Intelligent …, 2008

2022

In many board games and other abstract games, patterns have been used as features that can guide automated game-playing agents. Such patterns or features often represent particular configurations of pieces, empty positions, etc., which may be relevant for a game's strategies. Their use has been particularly prevalent in the game of Go, but also many other games used as benchmarks for AI research. Simple, linear policies of such features are unlikely to produce state-of-the-art playing strength like the deep neural networks that have been more commonly used in recent years do. However, they typically require significantly fewer resources to train, which is paramount for large-scale studies of hundreds to thousands of distinct games. In this paper, we formulate a design and efficient implementation of spatial state-action features for general games. These are patterns that can be trained to incentivise or disincentivise actions based on whether or not they match variables of the s...