Stanfordã®æ©æ¢°å¦ç¿ã³ã¼ã¹ã¯åå¦è ã«ãã³ããã¹ã¹ã¡ã§ãã
10æãã3ã¶æéåè¬ãã¦ãããã¹ã¿ã³ãã©ã¼ã大å¦ã®æ©æ¢°å¦ç¿ã®ãªã¼ãã³ã³ã¼ã¹ã§ããhttp://www.ml-class.org/ ã®ã³ã¼ã¹ãä¿®äºãã¾ããã
æ©æ¢°å¦ç¿ã®åå¦è ã¨ãã¦åè¬ãã¦ã¨ã¦ãæºè¶³æï¼å å®æããã£ãã®ã§ã·ã§ã¢ããã¦ãããã¾ãã
æ¥æ(1æã¹ã¿ã¼ã)ã同コースが開催されますã®ã§èå³ãããããæ¹ã¯ãã²ã©ããã
å¦ãã ãã¨
- æ師ããå¦ç¿ã
- ç·å½¢å帰ããã¸ã¹ãã£ãã¯å帰ããã¥ã¼ã©ã«ãããã¯ã¼ã¯ãSVM
- æ師æãå¦ç¿ã¢ã«ã´ãªãºã ã®ãã¥ã¼ãã³ã°ææ³
- Bias/Varianceåæãã¨ã©ã¼è§£æãå¦ç¿æ²ç·ã«ããåæãé©åç(precision)/åç¾ç(recall)/Få¤ã«ããåæ
- æ師ãªãå¦ç¿
- K-meansãã¢ãã¼ããªæ¤åºã主æååæ
- å¿ç¨
- ãã¥ã¼ã©ã«ãããã¯ã¼ã¯ã«ããæåèªèã·ã¹ãã ã®å®è£
- å調ãã£ã«ã¿ãªã³ã°ã«ãããªã³ã¡ã³ãã·ã¹ãã ã®å®è£
æ·å± ã¯é«ããªãã§ã
ãã®ã¯ã©ã¹ã¯é常ã®stanford open courseãããé£æ度çã«ã¯ä½ãè¨å®ããã¦ã¦ã大å¦ã¬ãã«ã®æ°å¦ãç¥ããªã人ããã©ã¤ã§ãããããã«éæã«å¿ è¦ãªåºç¤ç¥èã®è§£èª¬ãåºã¦ããã®ã§ãæ°è»½ã«åå ã§ãã¾ããï¼ãããªã®ä¸ã§ãããæ°å¦çãªç´°ãããã¨ã¯ç¥ããªãã¦ã大ä¸å¤«ã ãããã¨ãè¨ã£ã¦ã¾ãï½ï¼
ããããã¿ããåè¬ã«éãã¦ã®å¶ç´ãä¸åãªããæ¥ããã®æã¾ããå»ããã®è¿½ããã
ãããã£ãæå³ã§ãã¨ã¦ããªã¼ãã³ãªã³ã¼ã¹ãªã®ã§æ°è»½ã«åå ã§ãã¾ãã
ã¾ããã³ã¼ã¹ã2種é¡ã«åããã¦ãã¦ãèªåã®å¥½ããªã»ãã§åè¬ã§ããã¨ããã®ããã¤ã³ãã§ããï¼ã¡ãªã¿ã«åã¯Advanced Trackã§åè¬ãã¾ãã)
- Basic Track:ãããªè¬ç¾©ããããã¹ã(Review Question)ãããã°ã©ã æ¼ç¿ãã¹ã¦åè¬å¯è½ã ããæåºããªãã¦ãããã³ã¼ã¹ï¼è¦ãã ããããã ãï¼
- Advanced Track:ãããªè¬ç¾©ããããã¹ã(Review Question)ãããã°ã©ã æ¼ç¿ãã¹ã¦ãæåºããªããã°ãããªãã³ã¼ã¹ï¼è¦ãªããã ãããããªããã ãï¼
ããã§ãã¦çµæ§å®è·µçã§ã
ãã¨ãã£ã¦æ¦è«ãåºç¤ã ãã«ã¨ã©ã¾ãããããªãå®è·µãæèããè¬ç¾©ï¼æ¼ç¿å 容ã«ãªã£ã¦ãã¾ãã
- ãã¥ã¼ã©ã«ãããã¯ã¼ã¯ã«ããæåèªè
- å調ãã£ã«ã¿ãªã³ã°ã«ããæ ç»ã®ãªã³ã¡ã³ãã·ã¹ãã
ã¨ãã£ããããªãå®è·µçãªé¡æãåãä¸ãã¦ããã¦ãã¾ããè¬ç¾©ã§ããæ©æ¢°å¦ç¿ã®ã¨ã³ã¸ãã¢ã¨ãã¦ã®äººçã«å½¹ã«ç«ã¤ã¨æãã¾ããçãªè¨èãããã¾ããã
ã¾ããæ©æ¢°å¦ç¿ã¢ã«ã´ãªãºã ãå®éã«é©ç¨ããã¨ãã«æ©ããããªåé¡
- ã¢ã³ãã¼ãã£ããã»ãªã¼ãã¼ãã£ããã®åé¡
- åã¢ã«ã´ãªãºã ã®å©ç¹ã»æ¬ ç¹ï¼è¨ç®éçãªãã®ãå«ãã¦ï¼
- åæ£ç°å¢ã§ã®æ©æ¢°å¦ç¿ã¢ã«ã´ãªãºã ã®å®è£ Tips
ãªããã¸ã®å¯¾çãªãããå«ã¾ãã¦ãã¦ããã®ã³ã¼ã¹ãçµããã¨æ©æ¢°å¦ç¿ã®æè¡è
ã¨ãã¦éç¨ããããããªãããã¨ãã£ãæ°åã«ãªãã¾ãã
ãã£ãããã¼ã¹ã§å¦ã¹ã¾ã
å人çã«ã¯ããªãããããã¼ã¹ã§ååããªããã¾ããã®ã§ãæ¥å¤ä»äºã«è¿½ãããã¾ãæéãåããªãã¨ãããã社ä¼äººã®æ¹ã ã§ã大ä¸å¤«ã ã¨æãã¾ããæè¡è ã¨ãã¦ä»äºã®åéãã¬ã£ã¦ãã¤ãã¼ã¹ã«ã¹ãã«ã¢ããã§ããã®ã§ãã¨ã£ã¦ããªããã¡ã§ãã
ãã¼ã¹ã¨ãã¦ã¯ãããªæãããï¼
- å¹³æ¥ã®éå¤æéï¼çé1æéå¼±ï¼ã«ãæºå¡é»è»å ã§ã¬ã¡ã¤ãå¸ã確ä¿ãã¦iPadã§ãããªè¬ç¾©ãèãï¼Review Quizã«çãã
- åæ¥ã®ã©ã£ã¡ãã§æ°æéããã°ã©ãã³ã°æ¼ç¿(by octave)ã«å²ãã
ãâæ¼ç¿çµæã¯githubã§å ¬éä¸ï¼https://github.com/everpeace/ml-class-assignments
ããã°ã©ãã³ã°æ¼ç¿ã®å¤ãã¯ç©´åãå½¢å¼ã«ãªã£ã¦ãã¦ãã¢ã«ã´ãªãºã å ¨ä½ãå®è£ ããã®ã§ã¯ãªããè¬ç¾©ã§å¦ç¿ããã³ã¢ãªé¨åã®ããã°ã©ã ãåããã¨ãã£ãå½¢å¼ã§ãã
ãã¨ãã°ãã¥ã¼ã©ã«ãããã¯ã¼ã¯ã®æ¼ç¿ã§ã¯
- ãã¥ã¼ã©ã«ãããã¯ã¼ã¯ã®åºåé¨åã®è¨ç®
- ããã¯ãããã²ã¼ã·ã§ã³ã«ãããã©ã¡ã¼ã¿æ´æ°å¦ç
ã¨ãªã£ã¦ãã¾ããã
ãªã®ã§ãããã°ã©ãã³ã°æ¼ç¿ã®è² è·ã¯ããã¾ã§é«ããªãã«ãããããããå®è·µçãªæ©æ¢°å¦ç¿ã¢ã«ã´ãªãºã ã®å
¨å®¹ãç解ãããã¨ãã§ããããã«ãªã£ã¦ããã¨ããç¹ãã¨ã¦ããã¤ã³ãã«ãªãã¨æãã¾ãã
ã¿ããªã§å¦ã³ããã¾ã
ã¤ã³ã¿ã¼ãããã§åè¬ã¨ãããã¨ã§ãåè¬çã®æ°ãããªãå¤ãåé¢ãã¹ã¿ããã¯æ°äººã§ãã
ã¨ãããã¨ã§ãã¹ã¿ããã«ç´æ¥è³ªåã¯ã§ããªãã®ã§ãããã³ã¼ã¹å ã®ãã¼ã¸ã«ã¯åè¬çã©ãããããåãã§ããï¼ãã¡ããã¹ã¿ããã¨ãï¼Q&Aãã©ã¼ã©ã ããã£ã¦ããªãæ´»çºã«è³ªçãããã¦ãã¾ããã
ç§ãæ¼ç¿åé¡ã§è©°ã¾ã£ãæããããã¯èª²é¡ã®æ¹ãããããããããªããï¼ãªãã¦çåããã©ã¼ã©ã ã«æ¸ãè¾¼ãã ã¨ãã¯æ°åã§èª°ãããè¿çãæ¥ã¦ã³ã£ããã§ãããå ¨ä¸çã§åè¬çãããã®ã§24æéãã¤ã ã£ã¦ãã©ããã§ã ããã¯åãåå¼·ãã¦ããã§ããã
ãã®ç¹ã§ããä¸äººã§é²ãã¦ãã¦æ«æããããªãã¦å±éºæ§ã«ãèæ ®ãã¦ããã¦ããã®ããã¤ã³ãã§ãã
åèã§ãããä»åã®ã³ã¼ã¹ã®åè¬çåå¸ã表ããã¼ãããã(Globalç)ã§ãã
æ¥æ¬ãããªãèãã®ãã¡ãã£ã¨æ®å¿µã§ããã
- Global: http://i.imgur.com/rZ2yD.png
- APJ: http://i.imgur.com/lR8MK.png
- America: http://i.imgur.com/TJlyH.png
- Europe: http://i.imgur.com/ZmD4s.png
æ©æ¢°å¦ç¿ä»¥å¤ãè²ã ãã£ã¦ãã¹ã¹ã¡ã§ã
ä»æ(2011/10ã2011/12)ã¯äººå·¥ç¥è½ãæ©æ¢°å¦ç¿ããã¼ã¿ãã¼ã¹ãããã®ã³ã¼ã¹ã ã£ããã§ãããæ¥æï¼1æã¹ã¿ã¼ãï¼ã¯ãã¾ãã¾ãªåéã®ã³ã¼ã¹ãéå¬ãããããã§ãããã¡ãã機械学習のコースも再度開催されますã
ä»æã®æ©æ¢°å¦ç¿ãã¼ã¸ http://www.ml-class.org/ ã«æ¥æéå¬ãããã³ã¼ã¹ç¾¤ã¸ã®ãªã³ã¯ãããã¾ãã®ã§èå³ãããã°ãã²ã
åã®ç´ç·ã«è§¦ããã®ã¯
ã»Software as a Service
ã»Design and Analysis of Algorithms I
ã»Game Theory
ã»Cryptography
ã¨ãã£ãã¨ãããåã¯æ¥æã¯cryptographyãgame theoryãæ©ã¿ä¸ï¼ä¸¡æ¹ã¯ã¡ãã£ã¨é«è² è·ã«ãªã£ã¡ãããããªã®ã§ï¼ããã
å¦ãã ãã¡ããã¨ã¢ãã«ããã
æ代ã¯ãããã°ãã¼ã¿å ¨çæãã¯ã©ã¦ããåæ£è¨ç®ã大ãã£ã¼ãã¼ãwinnyã®ä½è ã®æ¹ãç¡ç½ªç¢ºå®ã§P2Pã®æè¡éçºã¯ã»ãå©ç½ªã«å½ãããªãå¤ä¾ãã§ãã¦æ¬å½ã«ããã£ãã¨ãã£ãæãã§ã
- ããã°ãã¼ã¿ãããã«æ´»ç¨ãããï¼âåæï¼ãã¤ã¤ãã³ã°ï¼
- ããã°ãã¼ã¿ãããã«é«éã«å¦çãããï¼âåæ£ï¼ã¹ã±ã¼ã«
ã¯ã¨ã£ã¦ãããããã¯ãã¼ã«ãªãã¨æããã¾ãã
ãããªæã«æ©æ¢°å¦ç¿ã®ã³ã¼ã¹ãåè¬ã§ããã®ã¯ã¨ã£ã¦ããããããã£ãã§ãããããä»äºã«ããã©ã¤ãã¼ãã«ãã¾ã£ããæ¯éãããããåå¼·ã§ããã®ã¯ã¨ã¦ãããããã§ãã
ãã ãä»åã®ããã°ã©ã æ¼ç¿ã§ä½¿ç¨ããoctaveã¯å®ç¨¼åã®æ©æ¢°å¦ç¿ã·ã¹ãã ã§ã¯ãã¾ãé©ç¨ããã¦ããªã模æ§ãã¨ãªãã¨ãã£ã±ãå®ç¨¼åã§ãéç¨ããè¨èªã§æ©æ¢°å¦ç¿ãå®è£ ã§ããæè¡ãã»ããã¨ããã§ãã
ã³ã¼ã¹å ã®ãã©ã¼ã©ã ã§ã¯ãã£ã±C/C++ã§ãããpythonã§ããããã¼ã¿ã®åå¦çãªãperlã ããã¨è²ã ãã人ããã¾ããã
octaveã¨ãã¢ã§è¯ãèãããRã§ãããRã¯å®ç¨¼åã·ã¹ãã ã§å®ç¸¾ã¯ããã®ã§ãããããRIHPEãããããããããã¯ä½¿ããã¦ããããã«ãªãã®ããªï¼æè¿ã ã¨Webç³»ã®ä¼ç¤¾ã§ã¯ãRã®ç¥èãããã¨æã¾ãããã¿ãããªæ±äººããããããªã®ã§ãç¾å ´ã§ä½¿ããã¦ããã®ããããã¾ãããã
åãé¸æã§ããå®ç¨çãªè¨èªã¨ããã¨ãã£ã±ãJavaãScalaãã¨ãããã¨ã§ããã£ã¡ç³»ã§ã¡ããã¡ããè¦ãããã¦ãã¾ãã
Hadoop上で動く機械学習ライブラリMahoutãã¦ã©ããã¯ãã¦ãããã©ãããæ°ã¶æåãããªã模æ§ã
Sparkã£ã¦ããScalaãã¼ã¹ã®åæ£è¨ç®ã®ã©ã¤ãã©ãªãæè¿è§¦ã£ã¦ã¿ããã©ãè¡åè¨ç®ã¨ããã¾ã ã¾ã å å®ãã¦ãªãã¦ãããã«åæ£æ©æ¢°å¦ç¿ã¢ã«ã´ãªãºã ã§æ¥½ãããã£ã¦ãããããªãããã§ããåæ£è¨ç®ã ãæ軽ã«ããã¦ããªãããã¨æããã ãã©ãï¼åæ£ã®è¡åè¨ç®ããå®è£ ããªããã ã¨ã¡ãã£ã¨é¢åããæ©æ¢°å¦ç¿ã®ã¢ã«ã´ãªãºã ãåä½ããããã§ãããããï¼
åæ£ã«ã¯å¯¾å¿ãã¦ããªãããã©ãScalaã ã¨ãScalalaã£ã¦ããç·å½¢ä»£æ°ã©ã¤ãã©ãªãããã®ã§ãããã§æ¼ç¿ã®ããç´ãã§ããã¦ã¿ããããªããªãã¦æã£ãããã¦ãã¾ããè¦ãã¨ããè¡åæ¼ç®ã¯çµæ§Octaveã£ã½ãæ¸ãããã§ããã®ã§ããããããããªã£ã¦æã£ãããæè¿è¯ãèãã®ä¸¦åã³ã¬ã¯ã·ã§ã³ã¨ã対å¿ãã¦ãã®ããªããã
ã¨ãã£ãã¨ããã§å®è·µã«åããæ´»åã¯ã¾ã ã¾ã åæ»æ°å³ãªãã ããã©ãããªãã¨ãå®è·µã§ä½¿ããæ©æ¢°å¦ç¿ã®æè¡ã磨ãã¦ããããã¨ããã§ãã
ã¯ã©ã¦ãããåæ£ã®æ代ããã£ã¦ãã¦ãSIerã§éããã¦ãããã°ã©ããªãã¦å¤ããè·ã«ãã¶ããæ代ããã£ã¦ãããããªã¼ãã¦æããã¥ã¼ã¹ãèãããããä¸ç¥è¾ãæ代ã§ãããæ¥ã ç²¾é²ãããã¨ã§ããããªç¯©ï¼ãµããï¼ããã£ã¦ãã¦ãã¡ããã¨å¼ã£ãããããã«ãªããããã®ã§ããã
ããã¾ãã