An Improvement of the Method on the Strategic Placing of Stones Based on the Possible Omission Number
スポンサーリンク
概要
- 論文の詳細を見る
In most Go playing systems, fuseki, or strategic placing of stones in the opening and early middle game, is based mainly on pattern knowledge in the corner and on the edge of the board. We are studying methods to play fuseki without pattern knowledge. In the opening game of Go, evaluation components cannot be clearly identified. In previous work, we have developed a method to estimate the strength of groups by the possible omission number (PON) and to evaluate a position by the strength and the size of the groups. This paper proposes an improvement of the method. On a collection of choose-one-among-four problems, the improved method reaches a score of 2-dan. We analyze unsolved problems on hand of a number of examples where the method fails and derive necessary future work from the analysis.
- 一般社団法人情報処理学会の論文
- 2000-10-26
著者
関連論文
- 学習システムRLSにおける意識度の利用
- 再帰型学習システムRLSの枠組みおよびESPによる実行システムの開発
- KPV法に基づく前打着時評価の再利用
- Computer Go as a Model of Research Evaluation(Session 3)
- 囲碁における族と領域
- An Improvement of the Method on the Strategic Placing of Stones Based on the Possible Omission Number
- 時刻入りの節を利用する RLS の学習
- 失敗原因解析による複雑なゲーム盤上パタン概念の学習
- 宣言的知識の利用による RLS の拡張
- Tree-Pairグラフの構成
- Realization of a Boolean Function Using An Extended Threshold Logic-4-Realization of a 2nd-order Threshold Function
- Realization of a Boolean Function Using An Extended Threshold Logic-3-Disjunctive Decomposition of a Boolean Function into 2nd-order Complete Monotonic Functions
- Realization of a Boolean Function Using An Extended Threshold Logic-2-2nd-order Monotonicity of 2nd-order Threshold Function
- Realization of a Boolean Function Using An Extended Threshold Logic-1-Realization of a 2nd-order Threshold Function using An Extended Threshold Logic
- 2次しきい値関数を実現するある拡張されたしきい値論理
- リダンダント変数を導入した拡張しきい値論理による2次しきい値関数の合成問題