Difference between revisions of "Hisayori Noda"

From Chessprogramming wiki
Jump to: navigation, search
Line 25: Line 25:
 
Ziosoft's ''the end of genesis T.N.K.evolution turbo type D'' was the first Shogi engine using [[Yu Nasu|Yu Nasu's]] new [[Evaluation Function|evaluation function]]
 
Ziosoft's ''the end of genesis T.N.K.evolution turbo type D'' was the first Shogi engine using [[Yu Nasu|Yu Nasu's]] new [[Evaluation Function|evaluation function]]
 
based on a shallow [[Neural Networks|neural network]], dubbed [[NNUE]] (&#398;U&#1048;&#1048; Efficiently Updatable Neural Networks) <ref>[[Yu Nasu]] ('''2018'''). ''&#398;U&#1048;&#1048; Efficiently Updatable Neural-Network based Evaluation Functions for Computer Shogi''.  Ziosoft Computer Shogi Club, [https://github.com/ynasu87/nnue/blob/master/docs/nnue.pdf pdf], [https://www.apply.computer-shogi.org/wcsc28/appeal/the_end_of_genesis_T.N.K.evolution_turbo_type_D/nnue.pdf pdf] (Japanese with English abstract)</ref>,
 
based on a shallow [[Neural Networks|neural network]], dubbed [[NNUE]] (&#398;U&#1048;&#1048; Efficiently Updatable Neural Networks) <ref>[[Yu Nasu]] ('''2018'''). ''&#398;U&#1048;&#1048; Efficiently Updatable Neural-Network based Evaluation Functions for Computer Shogi''.  Ziosoft Computer Shogi Club, [https://github.com/ynasu87/nnue/blob/master/docs/nnue.pdf pdf], [https://www.apply.computer-shogi.org/wcsc28/appeal/the_end_of_genesis_T.N.K.evolution_turbo_type_D/nnue.pdf pdf] (Japanese with English abstract)</ref>,
−
further refined in [[YaneuraOu]] using a [[Stockfish]] based Shogi adaptation, as well in competing Shogi playing entities such as [[WCSC29]] runner-up [[Kristallweizen-kai|Kristallweizen]] <ref>[https://github.com/Tama4649/Kristallweizen/ GitHub - Tama4649/Kristallweizen: 第29回世界コンピュータ将棋選手権 準優勝のKristallweizenです。]</ref>,
+
further refined in [[YaneuraOu]] using a [[Stockfish]] based Shogi adaptation, as well in competing Shogi playing entities such as [[WCSC29]] runner-up [[Kristallweizen]] <ref>[https://github.com/Tama4649/Kristallweizen/ GitHub - Tama4649/Kristallweizen: 第29回世界コンピュータ将棋選手権 準優勝のKristallweizenです。]</ref>,
 
both apparently with superhuman [[AlphaZero]] strength <ref>[http://www.talkchess.com/forum3/viewtopic.php?f=2&t=72754 The Stockfish of shogi] by [[Larry Kaufman]], [[CCC]], January 07, 2020</ref>.
 
both apparently with superhuman [[AlphaZero]] strength <ref>[http://www.talkchess.com/forum3/viewtopic.php?f=2&t=72754 The Stockfish of shogi] by [[Larry Kaufman]], [[CCC]], January 07, 2020</ref>.
  

Revision as of 21:28, 5 August 2020

Home * People * Hisayori Noda

Hisayori Noda, (野田久順, Nodchip)
a Japanese mathematician, computer scientist, and computer Shogi and chess programmer, who introduced NNUE to Stockfish, yielding in Stockfish NNUE. While affiliated with the Tokyo Institute of Technology from 2002 until 2011 [1], his research focused on signal processing along with harmonic analysis [2].

Shogi

Hisayori Noda is author and co-author of various Shogi playing entities within the Tanuki family [3], and member of their corresponding development teams. In 2016, his first Shogi programming involvement was with Tanuki no Mori as member of the Tanuki no Mori Production Committee along with Jun Okabe, Takahiro Suzuki, and Masatoshi Hidaka [4]. Subsequent teams were called The Minstrel's Ballad: Tanuki's Reign with Yu Nasu and at times Akio Kono joining [5] [6], and most recently Ziosoft Computer Shogi Club, with the World Computer Shogi Championship entries of Shouten Gensou Knights of Tanuki (2017, derived from Yaneura King by Motohiro Isozaki aka Yaneurao) [7] [8], the end of genesis T.N.K.evolution turbo type D (2018) [9] [10] and the 29th World Computer Shogi Champion [11] YaneuraOu [12] [13]. Hisayori Noda's open source Shogi engine Ultimate Fantasy Alte Matanupon (究極幻想アルテマタヌポン) is based on technology from YaneuraOu [14].

NNUE

Ziosoft's the end of genesis T.N.K.evolution turbo type D was the first Shogi engine using Yu Nasu's new evaluation function based on a shallow neural network, dubbed NNUE (ƎUИИ Efficiently Updatable Neural Networks) [15], further refined in YaneuraOu using a Stockfish based Shogi adaptation, as well in competing Shogi playing entities such as WCSC29 runner-up Kristallweizen [16], both apparently with superhuman AlphaZero strength [17].

Stockfish NNUE

In recent years, the Shogi developers have borrowed many ideas from Stockfish, and adopted its search and infrastructure. With the new NNUE revolution, Hisayori Noda's honest motivation was to give something back to the Stockfish community [18]. In 2019, as a proof of concept, he incorporated NNUE into Stockfish 10, After support and announcements by Henk Drost in May 2020 [19] and subsequent enhancements, Stockfish NNUE was established and recognized. In summer 2020, with more people involved in testing and training, the computer chess community bursts out enthusiastically due to its rapidly raising playing strength with different networks trained using a mixture of supervised and reinforcement learning methods. Despite the approximately halved search speed, Stockfish NNUE seemingly became stronger than its original [20].

Selected Publications

[21] [22]

External Links

References

  1. ↑ Hisayori Noda | LinkedIn
  2. ↑ Hisayori Noda, Akinori Nishihara (2009). Fast and Accurate Generalized Harmonic Analysis and Its Parallel Computation by GPU. IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, Vol. 92-A
  3. ↑ Japanese raccoon dog from Wikipedia
  4. ↑ 第26回世界コンピュータ将棋選手権 参加チーム 26th World Computer Shogi Championship participating teams
  5. ↑ WCSC26 participation list by Reijer Grimbergen, SHOGI-L, April 08, 2016
  6. ↑ The 27th World Computer Shogi Championship: participant list by Takenobu Takizawa, SHOGI-L, April 17, 2017
  7. ↑ Hisayori Noda, Jun Okabe, Takahiro Suzuki, Masatoshi Hidaka, Yu Nasu, Akio Kono (2017). Shouten Gensou Knights of Tanuki Appeal Document. The Minstrel's Ballad: Tanuki's Reign, pdf (Japanese)
  8. ↑ 第27回世界コンピュータ将棋選手権 参加チーム wcsc27
  9. ↑ 第28回世界コンピュータ将棋選手権 wcsc28
  10. ↑ Installation instruction of shogi engines, May 28, 2019
  11. ↑ 第29回世界コンピュータ将棋選手権 wcsc29
  12. ↑ やねうら王 公式サイト | コンピューター将棋 やねうら王 公式サイト YaneuraOu Official Website by Motohiro Isozaki aka Yaneurao
  13. ↑ GitHub - yaneurao/YaneuraOu: YaneuraOu is the World's Strongest Shogi engine(AI player), WCSC29 1st winner, educational and USI compliant engine, hosted by Motohiro Isozaki aka Yaneurao
  14. ↑ GitHub - nodchip/tanuki-: shogi engine (AI player), stronger than Bonanza6 , educational and tiny code(about 2500 lines) , USI compliant engine , capable of being compiled by VC++2015
  15. ↑ Yu Nasu (2018). ƎUИИ Efficiently Updatable Neural-Network based Evaluation Functions for Computer Shogi. Ziosoft Computer Shogi Club, pdf, pdf (Japanese with English abstract)
  16. ↑ GitHub - Tama4649/Kristallweizen: 第29回世界コンピュータ将棋選手権 準優勝のKristallweizenです。
  17. ↑ The Stockfish of shogi by Larry Kaufman, CCC, January 07, 2020
  18. ↑ Stockfish NNUE – The Complete Guide, June 19, 2020 (Japanese and English)
  19. ↑ Stockfish NN release (NNUE) by Henk Drost, CCC, May 31, 2020
  20. ↑ Can the sardine! NNUE clobbers SF by Henk Drost, CCC, July 16, 2020
  21. ↑ 野田久順 研究業績一覧 | T2R2 東京工業大学リサーチリポジトリ
  22. ↑ dblp: Hisayori Noda
  23. ↑ Hungarian algorithm from Wikipedia

Up one level

Retrieved from "https://www.chessprogramming.org/index.php?title=Hisayori_Noda&oldid=20534"