[B! machinelearning] f99aqã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

f99aq id:f99aq

machinelearningã«é–¢ã™ã‚‹f99aqã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (5)

${{author_name}}$

{{{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}}

Rettyæµã€Ž2200ä¸‡ãƒ¦ãƒ¼ã‚¶ã‚’æ”¯ãˆã‚‹æ©Ÿæ¢°å¦ç¿’åŸºç›¤ã€ã®ä½œã‚Šæ–¹ - Qiita
ã¿ãªã•ã‚“ã€ã“ã‚“ã«ã¡ã¯ã€‚Retty CTO ã®æ¨½çŸ³ã§ã™ã€‚ ã“ã®è¨˜äº‹ã¯ Retty Advent Calendar 25æ—¥ç›®ã§ã™ã€‚ãƒ¡ãƒªãƒ¼ã‚¯ãƒªã‚¹ãƒžã‚¹ã€‚ æ˜¨æ—¥ã¯ @ttakeoka ã®ã€ŽMFIã«ã‚€ã‘ã¦Rettyã®å–ã‚Šçµ„ã¿ã€ã§ã—ãŸã€‚ ä»Šå¹´ã‚‚æ®‹ã‚Šã‚ãšã‹ã«ãªã‚Šã¾ã—ãŸã€‚ã„ã‹ãŒãŠéŽã”ã—ã§ã™ã‹ï¼Ÿ Retty ã¯ã“ã® ï¼‘ å¹´ã§ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ãŒã»ã¼å€å¢—ã—ã¾ã—ãŸã€‚ãã‚Œã«ã‚ˆã£ã¦ã€æƒ…å ±ç™ºä¿¡è€…ãŒå¢—ãˆã€Advent Calendar ã«å‚åŠ å‡ºæ¥ã‚‹ã‚ˆã†ã«ãªã‚Šã¾ã—ãŸã€‚ã¿ã‚“ãªæ¥½ã—ãã†ã«ã—ã¦ã„ã¦ã€ã†ã‚Œã—ã„ã§ã™ã€‚ Retty Inc. Advent Calendar 2016 - Qiita ã•ã¦ã€ä»Šå¹´æœ€å¾Œã® Retty Advent Calendar è¨˜äº‹ã‚’æ›¸ãã¨ã„ã†ã“ã¨ã§ã€ã¯ã˜ã‚ã¯ ï¼‘å¹´ã®ã¾ã¨ã‚çš„å†…å®¹ã«ã—ã‚ˆã†ã‹ã¨æ€ã„ã¾ã—ãŸãŒã€ãã‚Œã§ã¯å¹³å‡¡ã§é¢ç™½ãã‚ã‚Šã¾ã›ã‚“ã€‚ãã“ã§ã€ãƒã‚¿ã«ãªã‚Šãã†ãªãƒžãƒ‹ã‚¢ãƒƒã‚¯ãªæŠ€è¡“çš„è¨˜äº‹ã§ç· ã‚ããã‚ŠãŸã„ã¨æ€ã„ã¾ã™ã€‚
f99aq 2017/02/11
docker

machinelearning
ãƒªãƒ³ã‚¯
å¤šè…•ãƒãƒ³ãƒ‡ã‚£ãƒƒãƒˆã«ã‚ˆã‚‹è¡¨ç¤ºã‚³ãƒ³ãƒ†ãƒ³ãƒ„ã®æœ€é©åŒ– - ã‚¯ãƒƒã‚¯ãƒ‘ãƒƒãƒ‰é–‹ç™ºè€…ãƒ–ãƒã‚°
ã“ã‚“ã«ã¡ã¯ã€‚æŠ€è¡“éƒ¨æ¤œç´¢ã‚°ãƒ«ãƒ¼ãƒ—ã®åŽŸå³¶ã§ã™ã€‚ ä¸Šã®ç”»åƒã¯ã€ã‚¹ãƒžãƒ¼ãƒˆãƒ•ã‚©ãƒ³ï¼ˆãƒ–ãƒ©ã‚¦ã‚¶ç‰ˆï¼‰ã§è¦‹ãŸã‚¯ãƒƒã‚¯ãƒ‘ãƒƒãƒ‰ã®æ¤œç´¢çµæžœãƒšãƒ¼ã‚¸ã§ã™ã€‚ãƒ¬ã‚·ãƒ”ã ã‘ã§ãªãã€ãƒ‹ãƒ¥ãƒ¼ã‚¹ã‚‚è¡¨ç¤ºã•ã‚Œã¦ã„ã¾ã™ãã€‚çŒ®ç«‹ã‚„æŽ²ç¤ºæ¿ã®ã‚¹ãƒ¬ãƒƒãƒ‰ãªã©ãŒè¡¨ç¤ºã•ã‚Œã‚‹ã“ã¨ã‚‚ã‚ã‚Šã¾ã™ã€‚ ã‚¯ãƒƒã‚¯ãƒ‘ãƒƒãƒ‰ã§ã¯ã€æ¤œç´¢çµæžœãƒšãƒ¼ã‚¸ã«è¡¨ç¤ºã™ã‚‹ã‚³ãƒ³ãƒ†ãƒ³ãƒ„ã‚’ã‚¯ã‚¨ãƒªãªã©ã«å¿œã˜ã¦æœ€é©åŒ–ã—ã¦ã„ã¾ã™ã€‚æœ€é©åŒ–ã¯ã€è†¨å¤§ãªãƒã‚°ãƒ‡ãƒ¼ã‚¿ã¨æœ€æ–°ã®æ©Ÿæ¢°å¦ç¿’ã‚’ç”¨ã„ã‚‹ã“ã¨ã§ã€å®Ÿç¾ã—ã¦ã„ã¾ã™ã€‚ã“ã®ã‚¨ãƒ³ãƒˆãƒªã§ã¯ã€ã‚¯ãƒƒã‚¯ãƒ‘ãƒƒãƒ‰ã«ãŠã‘ã‚‹ã‚³ãƒ³ãƒ†ãƒ³ãƒ„æœ€é©åŒ–ã®è£å´ã‚’ç´¹ä»‹ã—ã¾ã™ã€‚ æœ€é©åŒ–ã®èƒŒæ™¯ ã‚¹ãƒžãƒ¼ãƒˆãƒ•ã‚©ãƒ³ã®æ™®åŠã«ä¼´ã£ã¦ã€ãƒ¦ãƒ¼ã‚¶ãŒåˆ©ç”¨ã™ã‚‹ãƒ—ãƒ©ãƒƒãƒˆãƒ•ã‚©ãƒ¼ãƒ ã¯ PC ã‹ã‚‰ãƒ¢ãƒã‚¤ãƒ«ã«ã‚·ãƒ•ãƒˆã—ã¤ã¤ã‚ã‚Šã¾ã™ã€‚ã‚¯ãƒƒã‚¯ãƒ‘ãƒƒãƒ‰ã«ãŠã‘ã‚‹ãƒ¢ãƒã‚¤ãƒ«åˆ©ç”¨è€…ã®å‰²åˆã‚‚ã€ã“ã“ 2 å¹´ã§ 10% ä»¥ä¸Šå¢—åŠ ã—ã¾ã—ãŸã€‚æœ€è¿‘ã§ã¯ã€60% ä»¥ä¸Šã®ãƒ¦ãƒ¼ã‚¶ãŒãƒ¢ãƒã‚¤ãƒ«ã‹ã‚‰ã‚¢ã‚¯ã‚»ã‚¹ã—ã¦ã„ã¾ã™ã€‚ ãƒ¦ãƒ¼ã‚¶ã®åˆ©ç”¨å½¢æ…‹ãŒå¤‰åŒ–ã™ã‚Œã°ã€æ¤œç´¢çµæžœãƒšãƒ¼ã‚¸ã‚‚ãã®å¤‰åŒ–ã«å¯¾
f99aq 2014/11/09
algorithm

machinelearning

cookpad

ui
ãƒªãƒ³ã‚¯
SmartNewsã‚’æ”¯ãˆã‚‹æ©Ÿæ¢°å¦ç¿’
ãƒ‹ãƒ¥ãƒ¼ã‚¹ã‚¢ãƒ—ãƒªSmartNews(https://www.smartnews.be/)ã®èƒŒæ™¯ã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã«ã¤ã„ã¦TokyoWebMining30th(http://tokyowebmining30.eventbrite.com/)ã§è©±ã•ã›ã¦ã„ãŸã ã„ãŸéš›ã®è³‡æ–™ã§ã™ã€‚ â€¢SmartNews iphoneç‰ˆ: https://itunes.apple.com/jp/app/id579581125 â€¢SmartNews Androidç‰ˆ https://play.google.com/store/apps/details?id=jp.gocro.smartnews.android â€¢SmartNewsé–‹ç™ºè€…ãƒ–ãƒã‚° http://developer.smartnews.be/blog/Read less
f99aq 2013/10/20
algorithm

machinelearning
ãƒªãƒ³ã‚¯
Gmailå„ªå…ˆãƒˆãƒ¬ã‚¤è«–æ–‡ãƒ¡ãƒ¢ - kisa12012ã®æ—¥è¨˜
å…ƒè«–æ–‡ â€œThe Learning Behind Gmail Priority Inboxâ€, Douglas Aberdeen, Ondrey Pacovsky, Andrew Slater, LCCC : NIPS 2010 Workshop on Learning on Cores, Clusters and Clouds. http://research.google.com/pubs/archive/36955.pdf Gmail Priority Inboxã«ã¯PAãŒåˆ©ç”¨ã•ã‚Œã¦ã„ã‚‹ã¨è©±é¡Œã«ãªã£ã¦ã„ã‚‹ã®ã§ï¼Œèªã‚“ã§ã¿ã¾ã—ãŸï¼Ž ç°¡å˜ã«ã¾ã¨ã‚ PA + transfer learning + logistic model ãƒ©ãƒ³ã‚ãƒ³ã‚°å¦ç¿’ã§ã¯ï¼ŒthresholdãŒéžå¸¸ã«é‡è¦ãªåƒãã‚’æŒã¤ Gmail Priority Inboxã¯ã‚ãªãŸã®ãƒ¡ãƒ¼ãƒ«å‡¦ç†ã®æ™‚é–“ã‚’6%çŸç¸®ã—ã¦ãã‚Œã¾ã™ 1.The
f99aq 2011/02/13
google

algorithm

machinelearning
ãƒªãƒ³ã‚¯
å¤©æ°—äºˆå ±ã‹ã‚‰æ©Ÿæ¢°å¦ç¿’ã€é‡‘èžå·¥å¦ã¾ã§ - DO++
ã‚‚ã†éšåˆ†çµŒã¡ã¾ã™ãŒï¼Œå…ˆæ—¥CompViewç§‹ã®å¦æ ¡ã¨ã„ã†ã®ã«è¡Œãï¼Œ2æ³Š3æ—¥ã¿ã£ã¡ã‚Šæ©Ÿä¼šå¦ç¿’ã‚’å‹‰å¼·ã—ã¦ãã¾ã—ãŸï¼Žè¬›å¸«é™£ã¯è±ªè¯ã§ã©ã®è©±ã‚‚é¢ç™½ã‹ã£ãŸã®ã§ã™ãŒç‰¹ã«Elad Hazanã«ã‚ˆã‚‹"Prediction in the dark: the multi-armed bandit probl em"ãŒéžå¸¸ã«é¢ç™½ã‹ã£ãŸã§ã™ï¼Ž ãã®è©±ã‚’èª¬æ˜Žã™ã‚‹ãŸã‚ã«ï¼Œã¾ãšç°¡å˜ãªãŒã‚‰é©šãã¹ãæ€§èƒ½ã‚’é”æˆã™ã‚‹ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚’ç´¹ä»‹ã—ã¾ã—ã‚‡ã†ï¼Ž è§£ããŸã„å•é¡Œã¯ï¼Œæ¯Žæ—¥ï¼Œæ¬¡ã®æ—¥ã®å¤©æ°—ãŒæ™´ã‚Œã‹é›¨ã‹ã‚’äºˆæƒ³ã™ã‚‹å•é¡Œã§ã™ï¼Žtæ—¥ç›®ãŒæ™´ã‚Œã®å ´åˆ y(t)=1, é›¨ã®å ´åˆ y(t)=0ã¨è¡¨ã™ã“ã¨ã«ã—ã¾ã—ã‚‡ã†ï¼Žtæ—¥ç›®ã«y(t+1)ã‚’äºˆæƒ³ã™ã‚‹ã‚ã‘ã§ã™ï¼Ž ã•ã¦ã€è‡ªåˆ†ã¯å¤©æ°—ã®å°‚é–€å®¶ã§ã¯ãªã„ã®ã§ï¼Œè‡ªåˆ†ã§äºˆå ±ã›ãšã«ï¼Œå°‚é–€å®¶ã«é ¼ã‚‹ã“ã¨ã«ã—ã¦ã¿ã¾ã™ï¼ŽMäººã®å¤©æ°—äºˆå ±å£«ãŒã„ã¦ï¼Œãã‚Œãžã‚ŒãŒç‹¬è‡ªã«æ¬¡ã®æ—¥ã®å¤©æ°—ã‚’äºˆæƒ³ã—ã¦ã„ã¾ã™ï¼Žiäººç›®ã®å¤©æ°—äºˆå ±å£«ã®tæ—¥ç›®ã®äºˆå ±ã‚’p(i,t)
f99aq 2009/11/16
algorithm

machinelearning
ãƒªãƒ³ã‚¯
1