Title:
PROGRAM FOR FINDING AREA FOR COUNTER-MOVE POINTS IN GAME OF GO
Document Type and Number:
WIPO Patent Application WO/2022/079760
Kind Code:
A1
Abstract:
This Go algorithm: uses a scenario function to take nouns as 361 move points on a Go board surface; governs match opponents' executed move points with L2; governs own executed move points (executed counter-move points) with L4; governs a situation status of all the executed move points with L3; and governs a stone removal status of move points with W4. Every time a match opponent inputs move point information, the Go algorithm runs two-dimensional analysis of all modalities of the board surface captured by a subject vector L3, and subjects same to harmony division into partial modalities of any one type among a finite number of types. Next, as pertains to L2 (a, b) belonging to a partial modality G(W) that is one partial modality among one or more partial modalities, L4 (C, d) that is a center-of-gravity point belonging outside of the partial modality G(W), and coordinates (x, y) that indicate a possible counter-move point belonging to the partial modality G(W), the Go algorithm: identifies such coordinates (x, y) that a board surface measure between the coordinates (a, b) and the coordinates (c, d) and a board surface measure between the coordinates (x, y) and the coordinates (c, d) will make the same value, and that the board surface measure between the coordinates (a, b) and the coordinates (x, y) will be a minimum; and selects said coordinates (x, y) as a new own counter-move point.
More Like This:
Inventors:
NEGORO FUMIO (JP)
Application Number:
PCT/JP2020/038473
Publication Date:
April 21, 2022
Filing Date:
October 12, 2020
Export Citation:
Assignee:
NEGORO FUMIO (JP)
International Classes:
A63F13/80
Foreign References:
| JP2002102534A | 2002-04-09 | |||
| US20080004096A1 | 2008-01-03 |
Other References:
PANG YUAN FENG, ITO TAKESHI: "Determination System of Terms for Supporting Learning Go", TRANSACTIONS OF THE INFORMATION PROCESSING SOCIETY OF JAPAN (IPSJ), vol. 59, no. 4, 25 April 2018 (2018-04-25), JP , pages 1286 - 1294, XP055937142, ISSN: 0387-5806
"Strongest Go AI AlphaGo Dismantling New Book: Deep Learning, Monte Carlo Tree Search, Its Mechanism from Reinforcement Learning Augmented Revised Edition", 17 July 2018, SHOSUISHA CO., LTD., JP, ISBN: 978-4-7981-5777-1, article OTSUKI, SATOSHI: "Chapter 1.4 Basics of Go AI, Subchapter 1.4.5", pages: 053 - 059, XP009537146
"Strongest Go AI AlphaGo Dismantling New Book: Deep Learning, Monte Carlo Tree Search, Its Mechanism from Reinforcement Learning Augmented Revised Edition", 17 July 2018, SHOSUISHA CO., LTD., JP, ISBN: 978-4-7981-5777-1, article OTSUKI, SATOSHI: "Chapter 1.4 Basics of Go AI, Subchapter 1.4.5", pages: 053 - 059, XP009537146
Attorney, Agent or Firm:
ASAHI PATENT FIRM (JP)
Download PDF:
Previous Patent: SEMICONDUCTOR MODULE
Next Patent: PARAMETER ADJUSTMENT DEVICE AND PARAMETER ADJUSTMENT METHOD
Next Patent: PARAMETER ADJUSTMENT DEVICE AND PARAMETER ADJUSTMENT METHOD