Complement Naive BayesãSVMããéããã¼ã¨ä¸»å¼µãã¦ããã®ã§ãSVMããªããæè¿ã¯éããªã£ã¦ããããããã¨ããäºãç´¹ä»ãã¦ã¿ãããè¿å¹´ã¯SVMãªã©ã®å¦ç¿ãé«éã«è¡ãã¨ããææ¡ãè¡ããã¦ãããå®è£ ãå ¬éããã¦ãããã®ãããããã®ä¸ã®ä¸ã¤ã«liblinearã¨ããæ©æ¢°å¦ç¿ã©ã¤ãã©ãªããããã©ã¤ãã©ãªåããæ¨æ¸¬ã§ããéããliblinearã§ã¯ã«ã¼ãã«ã使ããã¨ãåºæ¥ãªãããããããã®åé度ãéãã大è¦æ¨¡ãã¼ã¿ã«é©ç¨ã§ããã¨ããå©ç¹ãããã liblinearãä½ã£ã¦ããã®ã¯libsvmã¨åãç 究ã°ã«ã¼ãã§ãChih-Jen Linãããã¸ã§ã¯ããªã¼ãã¼ã§ããããã ãlibsvmã¯ããªãæåãªã©ã¤ãã©ãªã§ãliblinearã«ã¯ãããã£ãæå³ã§å®å¿æããããï¼liblinearã®æ¹ã¯å ¬éããã¦ãã°ããã¯å²ã¨ãã°ããã£ãããããã©ãï¼ liblinearã«ã¯L1-SVM, L
{{#tags}}- {{label}}
{{/tags}}