æ©æ¢°å¦ç¿çéã®æ å ±åéæ¹æ³
ããã«ã¡ã¯ï¼Machine Learning Advent Calendar (MLAC) 2013ã®14æ¥ç®ãæ
å½ãã¾ãï¼[twitter:@kisa12012]ã§ãï¼æ®æ®µã¯å士å¦çã¨ãã¦ï¼åå°ãæ¾æµªããªããæ©æ¢°å¦ç¿ã®ç 究ããã¦ã¾ãï¼ä»åã®è¨äºã¯ãã¹ãã³ã§å·çãã¦ãã¾ãï¼ç¾å°æé(EST)ã§ã®ç· åã¯å®ã£ãã®ã§ã»ã¼ãâ¦ã§ãããï¼
æ¬æ¥ã¯æ©æ¢°å¦ç¿ã®æè¡çãªå
容ã®è©±ã§ã¯ãªãï¼çè
ãå®è·µãã¦ããæ©æ¢°å¦ç¿é¢é£ã®æ
å ±åéæ¹æ³ã«ã¤ãã¦çºãã¾ã*1ï¼å¤§ããåãã¦ï¼å¦ä¼æ
å ±ã®ç®¡çã»è«ææ
å ±ã®åéã»ãã®ä»ã®ä¸ç¨®ã«ã¤ãã¦è¿°ã¹ããã¨æãã¾ãï¼ä»åã®ãããã¯ã®å¤ãã¯ä»ã®åéã«ãéç¨ãã話ã«ãªã£ã¦ãããã¨æãã¾ãï¼ä»ã®åéã®æ¹ãã©ã®ããã«æ
å ±åéããã¦ããã®ããæ°ã«ãªãã¨ããã§ãï¼
å¦ä¼æ å ±ã®ç®¡ç
- ã¾ãã¯å¦ä¼æ
å ±ã®ç®¡çã«ã¤ãã¦ã§ãï¼æ©æ¢°å¦ç¿ã«é¢é£ããã«ã³ãã¡ã¬ã³ã¹ã¯ï¼ç¹ã«è¿å¹´ä¹±ç«æ°å³ã§ï¼é常ã«æ²¢å±±ããã¾ãï¼å
¨ã¦ããã§ãã¯ãã¦åãã®ã¯ï¼æ¥µä¸é¨ã®è¶
人ãé¤ãé常ã®äººåçç©ã«ã¯ä¸å¯è½ãªã¿ã¹ã¯ã§ãï¼ä»£è¡¨çãªå¦ä¼ãå人çãªå°è±¡ã¨ã¨ãã«ï¼ä»¥ä¸ã«ç°¡åã«è¨è¼ããã¨*2ï¼
- ICML/NIPS/AISTATS/ECMLPKDD ã¯æ©æ¢°å¦ç¿ç³»ã®æå 端ã®è«æãéã¾ãã«ã³ãã¡ã¬ã³ã¹ï¼ææ°ã®ååãè¦ãããã«ï¼å¿ ãæ¡æè«æã®ãã§ãã¯ãå¿ è¦ï¼
- UAI/ICLR/COLT ã¯ä¼è°æ¯ã«ã¯ã»ãå¼·ãï¼æ©æ¢°å¦ç¿ã®ä¸ã®ä¸åéã®ã¿ã対象ã¨ãã¦ããäºãå¤ããï¼ç·ãã¦è³ªã¯é«ãããèªèº«ãé¢é£ããå¦ä¼ã¯ãã§ãã¯ãå¿ é ï¼
- KDD/IJCAI/AAAI/SDM/ICDM/WWW ã§ã¯ä¸ã®æ©æ¢°å¦ç¿ææ³ã®å®å¿ç¨ã¸ã®æ¡å¼µãè±å¯ï¼WSDM/CIKMãè¦éããªãï¼
- ACL/EMNLP/ICCV/ECCV/CVPR/ICASSP ã¯è¨èªå¦çã»ç»åå¦çã»é³å£°å¦çããããã®åéã«é©å¿ããææ³ã®ææ¡ãå¤æ°ï¼ããã§ã®ã¢ã¤ãã¢ãææ³ãæ©æ¢°å¦ç¿ç³»ã¸éè¼¸å ¥ããããã¨ãå¤ã ï¼
- VLDB/SIGMOD ãªã©ã®ãã¼ã¿ãã¼ã¹ç³»ã§ããã¾ã«æ©æ¢°å¦ç¿ç³»ã®è«æãã¡ãã»ãï¼
- JMLR/MLJ/TPAMIãªã©ï¼ã¸ã£ã¼ãã«ãå¤æ°ï¼
- ãããã¾ãï¼ã¨ã¦ãå ¨é¨ã¯è¿½ããããªãããï¼å¦ä¼ã®éå¬æ å ±ãç· åæ å ±ã®å¹ççãªç®¡çãå¿ é ã§ãï¼ä»¥åã¯è²ã 試ãã¦ãã¾ãããï¼ä¸å¹´ç¨ã§æ´æ°ãæ¢ã¾ããµã¼ãã¹ãå¤ãæç¶çãªç®¡çã«ã¯ä¸åããªäºãå¤ãã£ãã§ãï¼æè¿ã¯ï¼èªåã§Wikicfpã®ãªã¹ãã管çããæ¹æ³ã«è½ã¡çãã¦ãã¾ãï¼
- Wikicfp http://www.wikicfp.com/cfp/
- 主ã«å¦ä¼ã®ç· åæ å ±ç®¡çã«å©ç¨ãã¦ãã¾ãï¼è«ææ稿ã®ãããã©ã¤ã³ãå¹ççã«ç®¡çåºæ¥ãç¹ã§æ°ã«å ¥ã£ã¦ã¾ãï¼
- èªåã管çãããã«ã³ãã¡ã¬ã³ã¹ããªã¹ãã«ç»é²ããã¨ï¼è«ææç¨¿ç· ååãéå¬åã«ã¡ã¼ã«ãRSSã§é£çµ¡ããã¦ããã¾ãï¼
- ã¾ãï¼ã¢ã¼ã«ã¤ãã«éå»ã®ã«ã³ãã¡ã¬ã³ã¹ãä¿åãã¦ããã¨æ¬¡å¹´åº¦ã®æ å ±ãã¡ã¼ã«çã§é£çµ¡ãã¦ããã¾ãï¼æç³»å表示ãicalå½¢å¼ã§ã®ç»é²ãå¯ï¼
- çè ã管çãã¦ããWikicfpã¯こちらï¼ä¸ã§æããä¼è°+αã®ç®¡çã«å©ç¨ãã¦ãã¾ãï¼
è«æåé
- èªåã®ç 究ã¨è¯ãä¼¼ãæ¢åç 究ãå¶ç¶çºè¦ãã¦ãã¾ãï¼æ°ã¶æéã®åªåãç¡é§ã«ãªã£ãçµé¨ã®ããæ¹ããã«ã¡ã¯ããã«ã¡ã¯ï¼
- æ¢åç 究ãµã¼ãã¤ãå¹ççã«åºæ¥ãã°å¤§å¹ ãªæéã®ç¯ç´ã«ãªãã¾ãï¼èå³ã®ããè«æãç¶²ç¾ çã«æ¢ãããã®æ¹æ³ã¨ãã¦ï¼çè ãå®è·µãã¦ããæ¹æ³ã以ä¸ã«çºãã¾ãï¼
- Arxiv RSSè³¼èª http://arxiv.org/
- Arxivã«ã¯æ¯æ¥ã®ããã«æ°ããè«æãæ稿ããã¦ãã¾ãï¼æ¯æ¥æ稿ããã¦ããè«æã®ã¿ã¤ãã«ãæãèªã¿ããã ãã§ãï¼æè¿ã®ç 究ãã¬ã³ãããã¼ããã«ã¤ãããã¨ãå¯è½ã§ãï¼
- æè¿ã¯ã«ã³ãã¡ã¬ã³ã¹ããå ã«Arxivã«è«æãæ²è¼ããããã¨ãå¤ãããï¼Arxivè«æã®ãã§ãã¯ã¯å¿ é ã§ãï¼ä»å¹´ã®NIPSè«æã®30%ä½ã¯æ¡æåã«Arxivã«å ¬éããã¦ããã®ã§ã¯ãªããï¼ã¨ããã®ãå人çãªæè¦ã§ãï¼[è¦æ¤è¨¼]
- ã¿ã¤ãã«ãåºãã£ããã¢ãã¹ããæãèªã¿ï¼ã¢ãã¹ããå¼·çã«åºãã£ããå 容ãèªãä½ã®å¿æã¡ï¼ãããã®ãã£ã«ã¿ã§éè¦ãªè«æãè¦éãã¦ãã¾ã£ã¦ãï¼è¯ãè«æãªããã®ãã¡ã«ã³ãã¡ã¬ã³ã¹ãã¸ã£ã¼ãã«ã«è¼ãã¯ããªã®ã§ï¼precisionãåªå ãã¦ãã¾ãï¼
- æ©æ¢°å¦ç¿ã®å ´åã¯こちらï¼
- çè ã¯ï¼stat.ML (æ©æ¢°å¦ç¿) / cs.AI (人工ç¥è½) / cs.CE (è¨ç®æ©å·¥å¦) / cs.CL (è¨ç®æ©è¨èªå¦) / cs.GT (ã²ã¼ã çè«) ã®äºã¤ãè³¼èªãã¦ãã¾ãï¼
- ä¼æ¥è«æ
- å¤ãã®ä¼æ¥ç 究æã¯æ©æ¢°å¦ç¿é¢é£ã®è«æãã«ãã´ãªåããã¦ããããï¼RSSã§åå¾å¯è½ã§ãï¼ãã¡ããã¿ã¤ãã«ãæµãèªã¿ãã¦ããã¨ï¼åä¼æ¥ã®ãã¬ã³ããèå³å¯¾è±¡ãå°ãã¤ã¡ã¼ã¸åºæ¥ãã®ã§ãªã¹ã¹ã¡ã§ãï¼ä»¥ä¸ã代表ä¾ã§ãï¼
- Microsoft Research Publications http://research.microsoft.com/apps/catalog/default.aspx?t=publications&ra=47200
- Google Research Publications http://research.google.com/pubs/ArtificialIntelligenceandMachineLearning.html
- Yahoo! Labs http://labs.yahoo.com/publication/?area=machine-learning
- Google Scholar http://scholar.google.com/
- æ©æ¢°å¦ç¿ããã£ããå¦ãã§ããã®ã ããï¼è«æãµã¼ãã¤ã«ãæ©æ¢°å¦ç¿ã®åãå©ç¨ãã¦æ¥½ããããï¼ã¨ãããã¨ã§ï¼Google Scholarã®è«æãªã³ã¡ã³ãæ©è½ãå©ç¨ãã¦ãã¾ãï¼
- åºæ¬çã«ã¯ï¼èªåãããã¾ã§ã«å·çããè«æãåºã«èªãã¹ãè«æããªã³ã¡ã³ããã¦ããã¾ãï¼
- ã¾ãã¯ï¼Google Scholar Citationsã«èªèº«ã®æ¸ããè«æãç»é²ãã¾ãããï¼
- ããã¨ï¼My Updatesã«è«æããªã³ã¡ã³ãããã¾ãï¼
- å·çè«æã®ã¿ã§ãªãGoogle Scholarã®è«æã¢ã¯ã»ã¹ãã¼ã¿ãã¬ã³ã¡ã³ãã«åæ ããã¦ããæ°ããã¾ãï¼[è¦æ¤è¨¼]
- ãã®ããï¼å¦ç¿ãã¼ã¿ã¨ãã¦Google Scholarçµç±ã§é¢ç½ãã£ãè«æã®ã¯ãªãã¯ãã¼ã¿ãä¸ãã¦ããã¨ï¼ãããã«ä¼¼ãè«æãèªåçã«éãã¦ãããããã«ãªãâ¦â¦æ°ããã¾ãï¼æ°ã®ããããããã¾ããï¼
ãã®ä»æ å ±ã®åé
- ããã°ããã¥ã¼ã¹è¨äºã§ãåéãããæ©æ¢°å¦ç¿é¢é£æ å ±ã¯æ²¢å±±ããã¾ãï¼ãã¡ããçè ãå®è·µãã¦ããæ¹æ³ãç°¡åã«ç´¹ä»ãã¾ãï¼
- ã¿ã°æå®ã¯ã¦ãªããã¯ãã¼ã¯
- ã¯ã¦ãªããã¯ãã¼ã¯ã§ã¯ã¿ã°ããã¼ã¯ã¼ãã»ãã¯ãæ°ãæå®ãã¦ï¼è©²å½ããè¨äºãRSSã§å¼ãæãäºãå¯è½ã§ãï¼試しに「機械学習」タグの付いたエントリを全部RSSで引き抜いてみるとï¼æ©æ¢°å¦ç¿é¢é£ã®æ¥æ¬èªè¨äºãç°¡åã«è¦ã¤ããã¾ãï¼
- æ¥æ¬èªã®ããã°ã»ãã¥ã¼ã¹è¨äºãslideshareã®ãã¬ã¼ã³è³æã¯ãã®æ¹æ³ã§ã»ã¼ç¶²ç¾ åºæ¥ãå°è±¡ãæã£ã¦ãã¾ãï¼æå¤ãªãã¨ã«ãã¤ãºããã¾ãç¡ãä¸æ¥ãããã®åéãå¤ããªãããï¼é常ã«éå®ãã¦ãã¾ãï¼
- çè ã¯ä»ã«ããèªç¶è¨èªå¦çããªã©ã®ãã¼ã¯ã¼ãã§RSSç»é²ããã¦ãã¾ãï¼ä»ã«ããããã®ãã¼ã¯ã¼ãçãç¥ã£ã¦ããï¼è¦ã¤ããæ¹ãããã°ï¼æ¯éæ¯éæãã¦é ããã¨å¹¸ãã§ãï¼
- æµ·å¤ããã°ã®æ å ±ã¯ä¸ã®æ¹æ³ã§ã¯ãã¾ãéãããã¾ããï¼çè ããã§ãã¯ãã¦ããæµ·å¤ããã°ã®ä¸é¨ãæç²ãã¦ä»¥ä¸ã«ç´¹ä»ãã¾ãï¼çè ãè³¼èªãã¦ããããã°ã¯ä»ã«ãããããããã¾ããï¼å人çèå³ã«å¼·ãå½±é¿ãåãã¦ããããå²æãã¾ãï¼ããã°éã§ããããã®è¨äºãç´¹ä»ãããäºãå¤ãããï¼èªèº«ã®åé対象ãå°ããã¤åºãã¦ããã®ãè¯ããã¨æãã¾ãï¼
- Machine Learning (Theory) http://hunch.net/
- Vowpal Wabbitã®ä½è ã®ä¸äººï¼VWãã«ã³ãã¡ã¬ã³ã¹ã®ãã¥ã¼ããªã¢ã«è³æçã®æ å ±å¤ãï¼ICMLçã®éå¶ãStatæ å ±ã®å ¬éã«ãç©æ¥µçãªã®ã§é常ã«éå®ãã¦ã¾ãï¼
- Machined Learnings http://www.machinedlearnings.com/
- Microsoftã®Research/Engineerï¼å¤§è¦æ¨¡ãã¼ã¿è§£ææè¡ã®è©±é¡ãä¸å¿ï¼ã¹ã¿ã¼ãã¢ããã¨ã³ã¸ãã¢çãçµé¨ããã¦ããï¼è«æã®è§£èª¬ãè¶ ããå®ç¨è¦ç¹ããã®ã³ã¡ã³ããå¤ãï¼é常ã«åå¼·ã«ãªãã¾ãï¼å人çã«ã¯ï¼ãã¼ã¿ãµã¤ã¨ã³ãã£ã¹ãã¨ã¯ãã®ãããªäººãæãè¨èã§ããã¨æã£ã¦ãã¾ãï¼
- Iâm a bandit https://blogs.princeton.edu/imabandit/
- 確ççæé©åããã³ãã£ããã®è¨äºãå¤æ°ï¼Guest Postãå¤ãï¼ããããã®åéã®å°é家ã«ããæ©æ¢°å¦ç¿è¨äºãèªããã¨ãåºæ¥ã¾ãï¼
- è¬æ¼ãããª
- Videolecturesãå®çªã§ãï¼ICMLãCVPRã¯techtalk.tvã§å ¬éããã¦ãã¾ããï¼ã¹ã©ã¤ãã¨æ åã®ãããã³ã°ãã¨ãã¦ããªããããªãå¤ããªã©ï¼åé¡ãå¤ããããã¾ã好ãã§ã¯ããã¾ããï¼ã«ã³ãã¡ã¬ã³ã¹è¬æ¼ãè¦ãã®ãè¯ãã§ããï¼ãªã¹ã¹ã¡ãªã®ã¯MLSSãªã©ã®ãã¥ã¼ããªã¢ã«ã®æ åè³æã§ãï¼
- Machine Learning Summer School (MLSS), La Palma 2012 - VideoLectures - VideoLectures.NET http://videolectures.net/mlss2012_lapalma/
- Machine Learning Summer School (MLSS), Bordeaux 2011 - VideoLectures - VideoLectures.NET http://videolectures.net/mlss2011_bordeaux/
- Machine Learning Summer School (MLSS), Cambridge 2009 - VideoLectures - VideoLectures.NET http://videolectures.net/mlss09uk_cambridge/
- Videolectures以å¤ã«ãè¯ãè¬æ¼ãããªã¯æ§ã ãªå ´æã«è»¢ãã£ã¦ãã¾ãï¼ä¾ãã°ï¼Simon instituteã®Big Data Programã§ã¯ï¼æå 端ã®ç 究ææçºè¡¨ããã¥ã¼ããªã¢ã«è³æã大éã«è»¢ãã£ã¦ãã¾ãï¼
- Big Data Boot Campã®ãããªãå°ãçºããã ãã§ãï¼ç¸å½ã«åãã¤ãã¾ãï¼ãã®ãããªæ©æ¢°å¦ç¿ã¤ãã³ããå¤ãéå¬ããã¦ããæãè¦ãã¨ï¼æ¥æ¬ã¨ã®ç°å¢ã®éããæãç¥ãããã¾ããï¼è¬ç¾©è³æããã¥ã¼ããªã¢ã«æ åçã®æ¢ãæ¹ã¯æªã ããã¨ãã£ãæ¹æ³ãè¦ã¤ãããã¦ãããï¼ã¢ã³ãããé«ãå¼µãç¶ãããããªãã¨ããã®ãç¾ç¶ã§ãï¼ããè¯ãæ¹æ³ãããã°æãã¦é ããã°å¹¸ãã§ãï¼
- æ¸ç±
- ç¹å®ã®ãããã¯ã«é¢ããè«æãµã¼ãã¤ãä¸å¿ã«ã¾ã¨ãããã¦ããFoundations and Trends in Machine Learningã·ãªã¼ãºã¨ï¼æ©æ¢°å¦ç¿å ¥éæ¸ã¨ãã¦ä¾¿å©ãªMurphy本ï¼ä»åã®MLACã§ã紹介ããã¦ãã¾ãï¼ï¼ããµã¼ãã¤æ¸ç±ã¨ãã¦ã¯ãªã¹ã¹ã¡ã§ãï¼
çè
ãå®è·µãã¦ããæ
å ±åéæ¹æ³ã«ã¤ãã¦ï¼ããã¾ã§ã ãã ãã¨ã¾ã¨ãã¦ãã¾ããï¼æ¥ã
ãããã®æ
å ±ãéãã¦åºç¤ä½åãã¤ãã¤ã¤ï¼èªåã®ã¿ã¹ã¯ã«ç¹ããããã®åå¼·ã¨ãã¦ï¼è¿å¹´ã®Dè«ã®ãµã¼ãã¤é¨åãããã¯æ¸ç±ãèªãâå¼ç¨è«æã§éè¦ãªãã®ãããã¯ã¢ãããã¦èªãâãããã®ç¥èãåå°ã«æè¿ã®è«æã«å½ããï¼ã®ããã»ã¹ãåãã®ãæççµè·¯ã§ããã¨æãã¾ãï¼
ããããï¼ã»ï¼å¹´å¾ã«ã¯ï¼ä»ã¨ã¯ã¾ã£ããç°ãªãæ
å ±åéæ¹æ³ã«å¤ãã£ã¦ãããã®ã¨æ³åãã¾ãï¼æ¥ã
æ¹è¯ãéãã¦ããè¯ãæ¹æ³ãæ¢ç´¢ãã¦ããããã§ãï¼ã¨ããããããªè¨èã§æ¬æ¥ã®MLACãç· ãããã¦ããã ãã¾ãï¼
*1:NIPS2013åå å ±åãæ¸ãããã¨ãèããã®ã§ããï¼I'm a banditã®ããã°è¨äºã§çè ãéè¦ã ã¨æã£ããã¨ã®ååç¨åº¦ã¯æ¢ã«è§£èª¬ããã¦ãã¾ãã¾ããï¼ç¢ºççæé©åãè¿æ¥å¾é æ³ã¾ããã®ä»å¹´ã®ãã¬ã¼ã¯ã¹ã«ã¼ã¯ï¼ãã¡ãã®è¨äºãåç §ããã°å¤§éæã«ææ¡ã§ããã¨æãã¾ãï¼ãã以å¤ã®åéã§é¢ç½ãã¨æã£ã話ã«ã¤ãã¦ã¯å¥ã®æ©ä¼ãããã°å¾æ¥ç´¹ä»ãããã¨æãã¾ãï¼ NIPSèªã¿ä¼ãªã©ã§ããããï¼
*2:機械学習の「朱鷺の社Wiki」ã§ãæ©æ¢°å¦ç¿é¢é£ã®å¦ä¼æ å ±ãç°¡åã«çºãããã¦ãã¾ãï¼