1 / 14

Position Evaluation in Computer Go

Position Evaluation in Computer Go. Martin Muller. Dept. of Computer Science and Information Engineering National Dong Hwa University Reporter : Lo Jung-Yun. Outline. Introduction Exact Evaluation in Explorer Heuristic Territory Evaluation in Explorer Full-Board Evaluation

cyphers
Télécharger la présentation

Position Evaluation in Computer Go

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Position Evaluationin Computer Go Martin Muller Dept. of Computer Science and Information Engineering National Dong Hwa UniversityReporter:Lo Jung-Yun

  2. Outline • Introduction • Exact Evaluation in Explorer • Heuristic Territory Evaluation in Explorer • Full-Board Evaluation • Open Problem and Future Work • Project Progress

  3. Introduction • Describe the position evaluation methods used in Explorer • Two different approach • Modularized heuristic evaluation using the concept of zones • An emphasis on exact evaluation methods for several types of local situations

  4. Exact Evaluation in Explorer • Safe Territory • Both static rules and method based on local search for detecting groups of blocks is safe • Capturing Race (semeai) • Also by static rules • End Game Area

  5. Heuristic Territory Evaluation in Explorer • Distinguish several different types of points • Divider and potential divider • Zone • Safe zone • Potential zone • Threatened zone • Unused zone • Point outside the territory • Near points • Junction points • Far-away points

  6. Heuristic Territory Evaluation in Explorer • Divider and potential divider • Computed using pattern matching • Safe zone • Are surrounded only by stones and dividers, and the interior is strongly controlled by stones • Potential zone • Weaker than safe territory, either because part of the boundary is only a potential divider, or because the control over the inside is not strong enough • Threatened zone • Weaker than potential zone • Unused zone • Contain strong opponent groups

  7. Different Types of Points • Divider and Potential Divider • Using pattern matching

  8. Different Types of Points • Zone, potential zone, threatened zone, and unused zone

  9. Point Outside The Territory • Divider-adjusted distances

  10. Point Outside The Territory

  11. Full-Board Evaluation • Weight setting

  12. Open Problems and Future Work • Maximize the chance of winning, not for score

  13. Project progress A.確定明確的方向及題目 B.充實基本知識 C.規劃系統及收集資訊 D.系統實作 E.驗收及改善 F.論文撰寫

  14. Project progress • Function • Open *.sgf file (using Neko’s parser) • Put stones on the board • Running • Check capture • Some data structure • Stone string • Stone group

More Related