Deep Learning From The First Principles | Bogdan Penkovsky, PhD
Breaking news! Artificial intelligence is taking over the world. Or it is not? Here is what you need to know about a deeper concept of meta-learning. Meta-learning is learning about learning. Learning how to learn belongs here too. Ever wondered how machines defeated the best human Go player Lee Sedol in 2016? A historical moment for the game that was previously considered to be very tough. What i
Haskellã¨ä½™ä»£æ•°(Coalgebra) - æœæ—¥ãƒãƒƒãƒˆã€€æŠ€è¡“者ブãƒã‚°
ã“ã“ã§ã¯Haskellã®ä¸ç´šè€…å‘ã‘ã®ãƒˆãƒ”ックを簡å˜ã«å–り上ã’ãŸã„ã¨æ€ã„ã¾ã™ã€‚ 今回ã¯ä½™ä»£æ•°(Coalgebra)ã«ã¤ã„ã¦ã§ã™ã€‚Haskellを書ã„ã¦ã„ã‚‹ã¨ã€Žä½™ï¼ˆãªã‚“ã¨ã‹ï¼‰ã€ã¨ã„ã†è¨€è‘‰ã‚’ã¿ã‹ã‘ã‚‹ã“ã¨ãŒã‚ã‚Šã¾ã™ã€‚ã“ã‚Œã¯è‹±èªžã®æŽ¥é 辞 Co- ã®è¨³ã§ã€ã“ã“ã§ã¯ä»£æ•°(Algebra)ã®åŒå¯¾(Dual)ã¨ã„ã†æ„味ã§ä½™ä»£æ•°ã¨å‘¼ã°ã‚Œã¦ã„ã¾ã™ã€‚ ã•ã¦Haskellã‚„OCamlã®ãƒ‡ãƒ¼ã‚¿åž‹ã¯ä¸€èˆ¬ã«ä»£æ•°çš„データ型(Algebraic data type)ã¨å‘¼ã°ã‚Œã¾ã™ã€‚ã“ã®ãƒ‡ãƒ¼ã‚¿åž‹ã«ãƒ‘ターンマッãƒãƒ³ã‚°ã‚’åŠ ãˆã¦ã€ï¼ˆã‚ªãƒ–ジェクト指å‘言語ãªã©ã¨æ¯”較ã—ãŸã¨ãã®ï¼‰é–¢æ•°åž‹è¨€èªžã®ç‰¹å¾´ã¨ç´¹ä»‹ã•ã‚Œã‚‹ã“ã¨ã‚‚多ã„ã¨æ€ã„ã¾ã™ã€‚実ã¯ã“ã®ãƒ‡ãƒ¼ã‚¿åž‹ã«ã¯ã‚‚ã¨ã‚‚ã¨ä½™ãƒ‡ãƒ¼ã‚¿(Codata)ã¨å‘¼ã°ã‚Œã‚‹ã‚ˆã†ãªç›¸æ–¹ãŒã„ãŸã®ã§ã™ãŒã€äººé¡žãŒæ‰±ã†ã«ã¯æ—©ã™ãŽãŸã®ã‹ã€ãƒ‡ãƒ¼ã‚¿åž‹ã«æ¯”ã¹ã‚‹ã¨ä½™ãƒ‡ãƒ¼ã‚¿ã¯é•·ã‚‰ãå½±ã®å˜åœ¨ã§ã‚ã‚Šã¾ã—ãŸï¼ˆã¡ãªã¿ã«ä½™ãƒ‡ãƒ¼ã‚¿ã¯å£èªžä¸Šã¯ã‚³ãƒ‡ãƒ¼ã‚¿ã¨å‘¼ã¶ã‹ã‚‚
Lensã ã‘ã§ä½œã‚‹ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯
ã“ã‚Œã¯ã€FOLIO Advent calendar 2021 ã®15日目ã®è¨˜äº‹ã§ã™ã€‚ åœè«–を機械å¦ç¿’ã«å¿œç”¨ã™ã‚‹è©±é¡Œã®ä¸€ã¤ã¨ã—ã¦Lensã§å¾®åˆ†å¯èƒ½ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ã‚’実装ã™ã‚‹è©±ã‚’紹介ã—ãŸã„ã¨æ€ã„ã¾ã™ã€‚ã¨ã¯ã„ãˆåœè«–ãªã©æ°—ã«ã›ãšLensを使ã£ãŸãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’実装ã—ã¦ã„ãã¾ã™ã€‚å¦ç¿’モデルã€èª¤å·®é–¢æ•°ã€å¦ç¿’ä¿‚æ•°ãªã©ã®åŸºæœ¬çš„ãªæ§‹æˆè¦ç´ ãŒå…¨ã¦Lens(ParaLens)ã¨ã—ã¦å®Ÿè£…ã§ãる様åを楽ã—ã‚“ã§ã„ãŸã ã‘ã‚Œã°ã¨æ€ã£ã¦ã„ã¾ã™ã€‚ Lensã£ã¦ä½•ï¼Ÿ Lensã¯ã„ã‚ゆる getter 㨠setter を組ã¿åˆã‚ã›ãŸãƒ‡ãƒ¼ã‚¿æ§‹é€ ã§ã™ã€‚ã™ãªã‚ã¡åž‹sã®ãƒ‡ãƒ¼ã‚¿åž‹ã‹ã‚‰åž‹aã®å€¤ã‚’å–り出ã™getter s -> a ã¨ã€åž‹sã®ãƒ‡ãƒ¼ã‚¿åž‹ã‚’åž‹aã®å€¤ã§æ›´æ–°ã—ã¦æ–°ã—ã„åž‹sã®ãƒ‡ãƒ¼ã‚¿åž‹ã‚’作æˆã™ã‚‹setter (s, a) -> s ã‹ã‚‰æˆã£ã¦ã„ã¾ã™ã€‚
仕事ã§ä½¿ã†Haskell
TL;DR Haskellã¯ã„ã„ãžã€‚ãŸã 仕事ã§ä½¿ã†ãªã‚‰ãƒ“ルドã€ãƒ‡ãƒ—ãƒã‚¤å‘¨è¾ºã¯å·¥å¤«ã™ã‚‹å¿…è¦ãŒè‰²ã€…出ã¦ãã‚‹ãžã€‚ ã‚ã€nixã‚‚ã„ã„ãžã€‚ ã¯ã˜ã‚ã« Haskellを用ã„ã¦ã„る会社HERPã«è»¢è·ã—ã¦ã‹ã‚‰ãã‚ãã‚1å¹´ãŒçµŒã¤ã®ã§ä¹…ã—ã¶ã‚Šã«è¨˜äº‹ã§ã‚‚書ã„ã¦ã¿ã¾ã™ã€‚ãã†ã„ãˆã°zennã§ã¯åˆæŠ•ç¨¿ã§ã™ã。 最近ã¯HERPã§Haskellを書ãã¤ã¤ã‚·ã‚¹ãƒ†ãƒ 基盤整備ã¿ãŸã„ãªã“ã¨ã‚’ã‚„ã£ã¦ã„ã¾ã™ã€‚ã‚ã¨ãƒžãƒãƒ¼ã‚¸ãƒ¡ãƒ³ãƒˆã¿ãŸã„ãªã“ã¨ã‚‚ã‚„ã£ã¦ãŸã‚Šã—ã¾ã™ã€‚ 僕ã®æ›¸ã„ãŸHaskell microserviceã¯æ—¢ã«ç¨¼åƒã—ã¦å£²ã‚Šä¸Šã’ã«è²¢çŒ®ã—ã¦ã„ã¾ã™ã€‚ ã‚ã€HERPã¯Haskell Foundationã®Functorスãƒãƒ³ã‚µãƒ¼ã«ãªã£ã¦ã„ã¾ã™ã€‚スãƒãƒ³ã‚µãƒ¼ã®åå‰ãŒé¢ç™½ã„ã§ã™ã‚ˆã。 Haskellを仕事ã§ä½¿ã†æ„Ÿæƒ³ 最高ã§ã™ã。簡å˜ä¾¿åˆ©å分速ã„ä¿å®ˆæ¥½æ‹¡å¼µæ¥½ã€ã¨ã„ã†æ„Ÿã˜ã§ã™ã€‚ 並行プãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ãƒ„ールã¨ã—ã¦stmãŒæä¾›ã•ã‚Œã¦ã„ã‚‹ã®ãŒãŠæ°—ã«å…¥ã‚Šã§
1
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
{{#tags}}- {{label}}
{{/tags}}