僕ã¯ã‚‚ã†ã€ãã‚“ãªã«ãƒŠã‚¤ãƒ¼ãƒ–ã˜ã‚ƒãªã„ã‚“ã - QiitaMachine Learning Advent Calendarã®20日目ã§ã™ã€‚ ã¯ã˜ã‚ã« Naive Bayes(NB)ã¨ãã®æ”¹è‰¯ç‰ˆã®Transformed Weight-normalized Complement Naive Bayes(TWCNB)ã€Averaged One-Dependence Estimators（AODE）ã¨ã„ã†æ‰‹æ³•ã«ã¤ã„ã¦è§£èª¬ã¨å®Ÿè£…を書ãã¾ã™ã€‚ Naive Bayes NBã¯ãƒ™ã‚¤ã‚ºã®å®šç†ã¨ç‰¹å¾´å¤‰æ•°é–“ã®ç‹¬ç«‹æ€§ä»®å®šã‚’用ã„ãŸåˆ†é¡žå™¨ã§ã™ã€‚文書ã®ã‚«ãƒ†ã‚´ãƒªåˆ†é¡žç‰ã§ã‚ˆã利用ã•ã‚Œã¾ã™ã€‚ NBã¯ã€äº‹ä¾‹$X$ã«å¯¾ã—$P(y|X)$ãŒæœ€å¤§ã¨ãªã‚‹ã‚¯ãƒ©ã‚¹$y$を分類çµæžœã¨ã—ã¦è¿”ã—ã¾ã™ã€‚$P(y|X)$ã¯ã€ãƒ™ã‚¤ã‚ºã®å®šç†ã‚’用ã„ã¦ã€ä»¥ä¸‹ã®ã‚ˆã†ã«å±•é–‹ãŒå¯èƒ½ã§ã™ã€‚ P(y|X) = \frac{P(y, X)}{P(X)} = \frac{P(X|y)*P(y)}{P(X)} \propto P
GitHub - herumi/prml
PRML副èªæœ¬ã€Œãƒ‘ターンèªè˜ã¨æ©Ÿæ¢°å¦ç¿’ã®å¦ç¿’ã€ã‚’出版ã—ã¾ã™ | TAKESAKO @ Yet another Cybozu Labs
{{#tags}}- {{label}}
{{/tags}}