Upgrade to Pro — share decks privately, control downloads, hide ads and more …

マイクロソフト発の麻雀AIの論文「Suphx: Mastering Mahjong with ...

lyakaap
February 25, 2025
2

マイクロソフト発の麻雀AIの論文「Suphx: Mastering Mahjong with Deep Reinforcement Learning」 について

マイクロソフト発の麻雀AIの論文「Suphx: Mastering Mahjong with Deep Reinforcement Learning」 についての解説

lyakaap

February 25, 2025
Tweet

Transcript

  1. ➭ך暴䗙ꆀ • 뒊ַ׸גְ׷晤 • سٓ • 䰍ג晤 • 䭯׍挿 •

     • MPPLBIFBEGFBUVSF如ل٦آד铡僇  ˟ 暴䗙ꆀך侧כⰋ鿇ד_⦐玎䏝
  2. 㷕统ך崧׸  ➂꟦ךفٖ؎ر٦ة׾⢪׏ג侄䌌֮׶㷕统 • 暴ח㢌׻׏׋ֿהכ׃גְזְךדⶴ䠥  ̔荈䊹㼎䨌ח״׷䓼⻉㷕统דׁ׵ח䓼ֻ • %JTUSJCVUFE3FJOGPSDFNFOU-FBSOJOH •

    1PMJDZHSBEJFOU • *NQPSUBODFTBNQMJOH • &OUSPQZSFHVMBSJ[BUJPO • (MPCBM3FXBSE1SFEJDUJPO • 0SBDMF(VJEJOH • 1BSBNFUSJD.POUF$BSMP1PMJDZ"EBQUBUJPOֿ׸搀׃ד⼧媮ז׏ג׷ 
  3. (MPCBM3FXBSE1SFEJDUJPOٌثك٦ءّٝ • 띾ꦾךٕ٦ٕ_ٓؐٝس㽷嶊⻉䖓ך剑穄أ؝،ד갫⡘寸㹀 • 㣓뒉ך㜥さծ갫⡘ח䘔ׄג㜠ꂹه؎ٝزָ䖤׵׸׷ • ⢽♧⡘ QU ✳⡘ QU

    ♲⡘QU 㔊⡘QU • 剑穄갫⡘ח״׷㜠ꂹ׾ؿ؍٦سغحؙה׃ג⢪欽׃׋㜥さ ♳䩛חفٖ؎דֹ׋ٓؐٝسה׉ֲדזְٓؐٝس׾⼒ⴽדֹזְ • ٓؐٝس׀הךأ؝،㟓幾׾ؿ؍٦سغحؙה׃׋㜥さ أ؝،׾幾׵ׅչ葺ְ،ؙءّٝպָ㶷㖈ׅ׷׋׭˟姻䔲ח،ؙءّٝך葺 ׃䝤׃׾鐰⣣דֹזְ ̔葺ְ䠬ׄךؿ؍٦سغحָؙדֹזְַ  ˟زحف׾然㹀ׇׁ׷׋׭ח剑穄ٓؐٝسד佦䠐ח䮶׶鴥׬ זו