[B! deep learning][machine learning] footnoteã®ãƒ

footnote id:footnote

deep learningã¨machine learningã«é–¢ã™ã‚‹footnoteã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (8)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

å›³ã¨ã‚³ãƒ¼ãƒ‰ã§å¿…ãšåˆ†ã‹ã‚‹ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®é€†ä¼æ’
é€£è¼‰ç›®æ¬¡ æœ¬ç¨¿ã¯ã€ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ï¼ˆä»¥ä¸‹ã€ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆï¼‰ã®ä»•çµ„ã¿ã‚„æŒ™å‹•ã‚’ã€æ•°å¦ç†è«–ã‹ã‚‰ã§ã¯ãªãPythonã‚³ãƒ¼ãƒ‰ã‹ã‚‰å¦ã¶ã“ã¨ã‚’ç›®æ¨™ã¨ã—ãŸé€£è¼‰ï¼ˆåŸºç¤Žç·¨ï¼‰ã®ç¬¬2å›žã§ã™ã€‚ã€Œé›£ã—ã„é«˜æ ¡ä»¥é™ã®æ•°å¦ã¯è‹¦æ‰‹ã ã‘ã©ã€ã‚³ãƒ¼ãƒ‰ã‚’èªã‚“ã§ãƒã‚¸ãƒƒã‚¯ã‚’ç†è§£ã™ã‚‹ã®ã¯å¾—æ„ï¼ã€ã¨ã„ã†æ–¹ã«ã¯ãƒ”ãƒƒã‚¿ãƒªã®è¨˜äº‹ã§ã™ã€‚ å‰å›žã®ç¬¬1å›žã§ã¯ã€æœ¬é€£è¼‰ã®ç›®çš„ã‚„ç‰¹å¾´ã‚’ç¤ºã—ãŸå¾Œã§ã€ã€Œãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆã®è¨“ç·´ï¼ˆå¦ç¿’ï¼‰å‡¦ç†ã‚’å®Ÿç¾ã™ã‚‹ãŸã‚ã«å¿…è¦ãªã“ã¨ã€ã¨ã—ã¦ã€ ã‚¹ãƒ†ãƒƒãƒ—ï¼ˆ1ï¼‰é †ä¼æ’ï¼š forward_prop()é–¢æ•°ã¨ã—ã¦å®Ÿè£…ï¼ˆå‰å›žï¼‰ ã‚¹ãƒ†ãƒƒãƒ—ï¼ˆ2ï¼‰é€†ä¼æ’ï¼š back_prop()é–¢æ•°ã¨ã—ã¦å®Ÿè£…ï¼ˆä»Šå›žï¼‰ ã‚¹ãƒ†ãƒƒãƒ—ï¼ˆ3ï¼‰ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ãƒ¼ï¼ˆé‡ã¿ã¨ãƒã‚¤ã‚¢ã‚¹ï¼‰ã®æ›´æ–°ï¼š update_params()é–¢æ•°ã¨ã—ã¦å®Ÿè£…ï¼ˆæ¬¡å›žï¼‰ã€‚ã“ã‚Œã«ã‚ˆã‚Šãƒ¢ãƒ‡ãƒ«ãŒæœ€é©åŒ–ã•ã‚Œã‚‹ ã¨ã„ã†3å¤§ã‚¹ãƒ†ãƒƒãƒ—ã‚’ç¤ºã—ã¾ã—ãŸã€‚å‰å›žã¯ã“ã®ã†ã¡ã®ã€Œã‚¹ãƒ†ãƒƒãƒ—ï¼ˆ1ï¼‰é †ä¼æ’ã€ã¾ã§å®Ÿè£…ãŒå®Œäº†ã—ã¦ã„ã¾ã™ã€‚ä»Šå›ž
footnote 2022/02/19
[

machine learning

deep learning

python
ãƒªãƒ³ã‚¯
ç¬¬1å›žã€€åˆã‚ã¦ã®ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯å®Ÿè£…ã€ã¾ãšã¯æº–å‚™ã‚’ã—ã‚ˆã† â€• ä»•çµ„ã¿ç†è§£Ã—åˆå®Ÿè£…ï¼ˆå‰ç·¨ï¼‰
ç¬¬1å›žã€€åˆã‚ã¦ã®ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯å®Ÿè£…ã€ã¾ãšã¯æº–å‚™ã‚’ã—ã‚ˆã† â€• ä»•çµ„ã¿ç†è§£Ã—åˆå®Ÿè£…ï¼ˆå‰ç·¨ï¼‰ï¼šTensorFlow 2ï¼‹Kerasï¼ˆtf.kerasï¼‰å…¥é–€ æ©Ÿæ¢°å¦ç¿’ã®å‹‰å¼·ã¯ã“ã“ã‹ã‚‰å§‹ã‚ã¦ã¿ã‚ˆã†ã€‚ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®åŸºç›¤æŠ€è¡“ã§ã‚ã‚‹ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ï¼ˆNNï¼‰ã‚’ã€çŸ¥è˜ã‚¼ãƒã®çŠ¶æ…‹ã‹ã‚‰æ¦‚ç•¥ã‚’æŠ¼ã•ãˆã¤ã¤å®Ÿè£…ã—ã¦ã¿ã‚ˆã†ã€‚ã¾ãšã¯ãƒ¯ãƒ¼ã‚¯ãƒ•ãƒãƒ¼ã‚’æ¦‚è¦³ã—ã¦ã€ãƒ‡ãƒ¼ã‚¿å›žã‚Šã®å‡¦ç†ã‹ã‚‰å§‹ã‚ã‚‹ã€‚
footnote 2022/02/10
machine learning

deep learning

study
ãƒªãƒ³ã‚¯
ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯å…¥é–€
Pythonã‚³ãƒ¼ãƒ‰ã§ç†è§£ã™ã‚‹ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯å…¥é–€ ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®ä»•çµ„ã¿ã‚„æŒ™å‹•ã‚’ã€æ•°å¦ç†è«–ã‹ã‚‰ã§ã¯ãªãã€Pythonã‚³ãƒ¼ãƒ‰ã‹ã‚‰ç†è§£ã—ã‚ˆã†ã€‚ãƒ•ãƒ«ã‚¹ã‚¯ãƒ©ãƒƒãƒã§ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ï¼ˆDNNï¼šDeep Neural Networkï¼‰ã‚’å®Ÿè£…ã—ã¦ã„ãã€‚ ç¬¬1å›žã€€Pythonã§ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’æ›¸ã„ã¦ã¿ã‚ˆã†ï¼ˆ2022/02/09ï¼‰ æœ¬é€£è¼‰ï¼ˆåŸºç¤Žç·¨ï¼‰ã®ç›®çš„ ãƒ»æœ¬é€£è¼‰ï¼ˆåŸºç¤Žç·¨ï¼‰ã®ç‰¹å¾´ ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®å›³ è¨“ç·´ï¼ˆå¦ç¿’ï¼‰å‡¦ç†å…¨ä½“ã®å®Ÿè£… ãƒ¢ãƒ‡ãƒ«ã®å®šç¾©ã¨ã€ä»®ã®è¨“ç·´ãƒ‡ãƒ¼ã‚¿ ã‚¹ãƒ†ãƒƒãƒ—1. é †ä¼æ’ã®å®Ÿè£… ãƒ»1ã¤ã®ãƒŽãƒ¼ãƒ‰ã«ãŠã‘ã‚‹é †ä¼æ’ã®å‡¦ç† ãƒ»é‡ã¿ä»˜ãç·šå½¢å’Œ ãƒ»æ´»æ€§åŒ–é–¢æ•°ï¼šã‚·ã‚°ãƒ¢ã‚¤ãƒ‰é–¢æ•° ãƒ»æ´»æ€§åŒ–é–¢æ•°ï¼šæ’ç‰é–¢æ•° ãƒ»é †ä¼æ’ã®å‡¦ç†å…¨ä½“ã®å®Ÿè£… ãƒ»é †ä¼æ’ã«ã‚ˆã‚‹äºˆæ¸¬ã®å®Ÿè¡Œä¾‹ ãƒ»ä»Šå¾Œã®ã‚¹ãƒ†ãƒƒãƒ—ã®æº–å‚™ï¼šé–¢æ•°ã¸ã®ä»®å¼•æ•°ã®è¿½åŠ ç¬¬2å›žã€€å›³ã¨ã‚³ãƒ¼ãƒ‰ã§å¿…ãšåˆ†ã‹ã‚‹ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã®é€†ä¼æ’ï¼ˆ2022/02/16ï¼‰
footnote 2022/02/09
machine learning

deep learning

neural network
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’å¦ã³å§‹ã‚ãŸæ–¹ã¸ã€€æ±äº¬å¤§å¦ï¼æ¾å°¾è±Šæ•™æŽˆã®å‹•ç”» - Qiita
Deleted articles cannot be recovered. Draft of this article would be also deleted. Are you sure you want to delete this article? #1.ã¯ã˜ã‚ã« ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’å¦ã³å§‹ã‚ãŸæ–¹ã«ã¨ã£ã¦ã€æ±äº¬å¤§å¦ï¼æ¾å°¾æ•™æŽˆã®å‹•ç”»ã‚’è¦‹ã‚‹ã“ã¨ã¯ã€ã¨ã¦ã‚‚åˆºæ¿€çš„ã§å‹‰å¼·ã«ãªã‚Šé¢ç™½ã„ã‚‚ã®ã ã¨æ€ã„ã¾ã™ã€‚ä»Šå›žã€æ¾å°¾æ•™æŽˆã®è¬›æ¼”ã«åŠ ãˆã¦å¯¾è«‡ã‚„ãƒ‘ãƒãƒ«ãƒ‡ã‚£ã‚¹ã‚«ãƒƒã‚·ãƒ§ãƒ³ã‚‚å«ã‚ãŸå‹•ç”»ã®ãƒªãƒ³ã‚¯ã‚’ã¾ã¨ã‚ã¾ã—ãŸã®ã§ã€ã‚ˆã‚ã—ã‹ã£ãŸã‚‰è¦‹ã¦ä¸‹ã•ã„ã€‚ ãŠã™ã™ã‚ã¯ã€01, 05, 14, 16, 23, 27 ã§ã™ã€‚ #2.å‹•ç”»ãƒªãƒ³ã‚¯ è¬›æ¼”ã«ã¯â˜…ã®è¡¨ç¤ºãŒã—ã¦ã‚ã‚Šã¾ã™ã€‚ â–¡2012å¹´ â˜…01.Computer will be more clever than human beings æ±äº¬å¤§å¦ç‰ˆTEDã§ã™ï¼ˆã‚‚ã¡ã‚ã‚“æ—¥æœ¬èªžã§ã™
footnote 2020/09/08
machine learning

deep learning

programming
ãƒªãƒ³ã‚¯
ã‚¼ãƒã‹ã‚‰å§‹ã‚ã‚‹æ·±å±¤å¼·åŒ–å¦ç¿’ï¼ˆNLP2018è¬›æ¼”è³‡æ–™ï¼‰/ Introduction of Deep Reinforcement Learning
Introduction of Deep Reinforcement Learning, which was presented at domestic NLP conference. è¨€èªžå‡¦ç†å¦ä¼šç¬¬24å›žå¹´æ¬¡å¤§ä¼š(NLP2018) ã§ã®è¬›æ¼”è³‡æ–™ã§ã™ã€‚ http://www.anlp.jp/nlp2018/#tutorialRead less
footnote 2018/03/26
machine learning

deep learning
ãƒªãƒ³ã‚¯
åˆå¿ƒè€…å‘ã‘ã€Œæ©Ÿæ¢°å¦ç¿’ã¨ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã®é•ã„ã€ã‚’ã‚·ãƒ³ãƒ—ãƒ«ã«è§£èª¬
By darkday AI(äººå·¥çŸ¥èƒ½)ãŒå¤§ããªè©±é¡Œã¨ãªã£ã¦ã„ã‚‹ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ãƒ¼ã‚µã‚¤ã‚¨ãƒ³ã‚¹ã®ä¸–ç•Œã§ã€ãã®æŠ€è¡“ã‚’æ”¯ãˆã¦ã„ã‚‹ã®ãŒã€Œãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã€ã§ã™ã€‚ä¸€æ–¹ã€ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ãƒ¼ã‚’ä½¿ã£ãŸã€Œæ©Ÿæ¢°å¦ç¿’ã€ã¨ã„ã†è¨€è‘‰ã‚’è€³ã«ã™ã‚‹ã“ã¨ã‚‚å¤šã„ã‚‚ã®ã§ã™ãŒã€å®Ÿã¯ãã®é•ã„ãŒã‚ˆãã‚ã‹ã‚‰ãªã„äººã‚‚å¤šã„ã¯ãšã€‚ãã‚“ãªä¸¡è€…ã®é•ã„ã‚’ã€æ•°å¦çš„è¨ˆç®—ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ã€ŒMATLABã€ã®é–‹ç™ºå…ƒã§ã‚ã‚‹MathWorksãŒç°¡å˜ã«è§£èª¬ã—ã¦ã„ã¾ã™ã€‚ Introduction to Deep Learning: Machine Learning vs Deep Learning - YouTube æ©Ÿæ¢°å¦ç¿’ã‚‚ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚‚ã€å¦ç¿’ãƒ¢ãƒ‡ãƒ«ã‚’æä¾›ã—ã¦ãƒ‡ãƒ¼ã‚¿ã‚’åˆ†é¡žã™ã‚‹ã“ã¨ã«ä½¿ã‚ã‚Œã‚‹æŠ€è¡“ã§ã™ã€‚ãã®åƒãã‚’è§£èª¬ã™ã‚‹ã®ã«ã‚ˆãç”¨ã„ã‚‰ã‚Œã‚‹ã®ãŒã€çŠ¬ã¨çŒ«ã®ç”»åƒã‚’åˆ†é¡žã™ã‚‹ã¨ã„ã†ä¾‹ã€‚ã“ã®ç”»åƒã®å ´åˆã€ã»ã¼å…¨ã¦ã®äººãŒå·¦ãŒçŠ¬ã€å³ãŒçŒ«ã¨ç”ãˆã‚‹ã¯ãšã€‚ ã—ã‹ã—ã€åˆ¥ã®ç”»åƒã‚’æŒã£ã¦ããŸæ™‚ã€ãã‚Œ
footnote 2017/08/05
machine learning

deep learning
ãƒªãƒ³ã‚¯
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’ç”¨ã„ã¦ã€Œå†™çœŸã®è¦‹ãŸç›®ã®ç‰¹å¾´ã€ã‚’åˆ¥ã®å†™çœŸã«è»¢é€ã—ã¦ã—ã¾ã†ã€ŒDeep Photo Style Transferã€
ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã‚’ç”¨ã„ã¦ãƒ™ãƒ¼ã‚¹ã¨ãªã‚‹ç”»åƒã«ä»–ã®ç”»åƒã®ã‚¹ã‚¿ã‚¤ãƒ«(è¦‹ãŸç›®ã®ç‰¹å¾´)ã‚’ä»˜ä¸Žã™ã‚‹ã“ã¨ã§ã€æ–°ã—ã„ç”»åƒã‚’ç”Ÿæˆã™ã‚‹ã“ã¨ãŒã§ãã‚‹ã€ŒDeep Photo Style Transferã€ãŒã€ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢é–‹ç™ºãƒ—ãƒã‚¸ã‚§ã‚¯ãƒˆã®å…±æœ‰ãƒ—ãƒ©ãƒƒãƒˆãƒ•ã‚©ãƒ¼ãƒ ã§ã‚ã‚‹GitHubä¸Šã§å…¬é–‹ã•ã‚Œã¦ã„ã¾ã™ã€‚ GitHub - luanfujun/deep-photo-styletransfer: Code and data for paper "Deep Photo Style Transfer" https://github.com/luanfujun/deep-photo-styletransfer ã“ã‚Œã¾ã§ã‚‚ã€ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ãªã©ã‚’ç”¨ã„ã¦ãƒ™ãƒ¼ã‚¹ã¨ãªã‚‹ç”»åƒã«ä»–ã®ç”»åƒã®ã‚¹ã‚¿ã‚¤ãƒ«ã‚’è»¢é€ã™ã‚‹ã€ã¨ã„ã†è©¦ã¿ãŒå˜åœ¨ã—ãŸãã†ã§ã™ã€‚ã—ã‹ã—ã€ãƒ™ãƒ¼ã‚¹ãƒ»ã‚¹ã‚¿ã‚¤ãƒ«ã®ä¸¡æ–¹ã«å†™çœŸã‚’ä½¿ã£ã¦ã‚‚ã€å‡ºåŠ›ã—ãŸç”»åƒãŒå†™çœŸã£ã½ããªã„ã€Œçµµã®ã‚ˆã†ãªã‚†ãŒã¿ã€ã‚’ã‚‚
footnote 2017/03/30
machine learning

deep learning

photo
ãƒªãƒ³ã‚¯
IBMã®äººå·¥çŸ¥èƒ½ã€Œãƒ¯ãƒˆã‚½ãƒ³ã€ã€åŒ»è€…ãŒæ€ã„ã‚‚ã‚ˆã‚‰ã¬æ²»ç™‚æ³•ã‚’ç¶šã€…ç™ºè¦‹ï¼ˆå°æž— é›…ä¸€ï¼‰ @gendai_biz
IBMãŒç¤¾é‹ã‚’ã‹ã‘ã¦å•†ç”¨åŒ–ã‚’é€²ã‚ã¦ã„ã‚‹AIã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã€Œãƒ¯ãƒˆã‚½ãƒ³ã€ãŒã€ã€ŒåŒ»ç™‚ã€ã‚’ã¯ã˜ã‚æ§˜ã€…ãªå¿œç”¨åˆ†é‡Žã§å®ŸåŠ›ã®ç‰‡é±—ã‚’è¦‹ã›å§‹ã‚ã¦ã„ã‚‹ã€‚ â—â€œIBM Is Counting on Its Bet on Watson, and Paying Big Money for Itâ€ã€€The New York Times, OCT.17, 2016 ä¸Šã®è¨˜äº‹ã«ã‚ˆã‚Œã°ã€æœ€è¿‘ç™ºè¡¨ã•ã‚ŒãŸIBMã®å››åŠæœŸåŽå…¥(revenue)ã¯17æœŸé€£ç¶šã§æ¸›å°‘ã—ãŸã‚‚ã®ã®ã€ã‚ˆã†ã‚„ãä¸‹ã’æ¢ã¾ã‚‹æ°—é…ã‚’ç¤ºã—ã¦ã„ã‚‹ã€‚ä¸ã§ã‚‚ãƒ¯ãƒˆã‚½ãƒ³ã®ã‚ˆã†ãªã€ŒAI(äººå·¥çŸ¥èƒ½)ã€ã‚„ã€Œãƒ‡ãƒ¼ã‚¿è§£æžã€ãªã©æ¬¡ä¸–ä»£ãƒ“ã‚¸ãƒã‚¹ãŒã€éžå¸¸ã«é«˜ã„ä¼¸ã³ã‚’è¨˜éŒ²ã—ã€åŒç¤¾ã®åŽç›Šæ”¹å–„ã«è²¢çŒ®ã—ã¦ã„ã‚‹ã¨ã„ã†ã€‚ åŠ›ã¥ãã§è¨€èªžèƒ½åŠ›ã‚’ç¿’å¾— ãƒ¯ãƒˆã‚½ãƒ³ã¯2011å¹´ã«ã€ç±³å›½ã®è‘—åãªãƒ†ãƒ¬ãƒ“ãƒ»ã‚¯ã‚¤ã‚ºç•ªçµ„ã€Œã‚¸ãƒ§ãƒ‘ãƒ‡ã‚£(Jeopardy)ã€ã«å‡ºæ¼”ã—ã¦ã€ãã®æ´ä»£ãƒãƒ£ãƒ³ãƒ”ã‚ªãƒ³(ã‚‚ã¡ã‚ã‚“äººé–“)ã‚’æ‰“ç ´ã™ã‚‹ã“ã¨ã§ä¸€èºæœ‰åã«ãª
footnote 2016/10/28
AI

machine learning

natural language processing

deep learning
ãƒªãƒ³ã‚¯
1