次ã¸: ã¯ããã« [ãã¥ã¼ããªã¢ã«è¬æ¼] ã«ã¼ãã«ãã·ã³ èµ¤ç© æ太é1 Shotaro Akaho s.akaho@aist.go.jp æ¦è¦: ãµãã¼ããã¯ã¿ãã·ã³ (SVM) ã«ä»£è¡¨ãããã«ã¼ãã«ãç¨ããå¦ç¿æ©æ¢°ã«ã¤ãã¦è§£èª¬ããï¼ ãããã«ã»ã¼å ±éãã¦ããã®ã¯ï¼åºæ¬çã«ç·å½¢ã®å¦ç¿æ©æ¢°ã®å»¶é·ç·ä¸ã«ããï¼ ãã¼ã«ã«ãããã ã®åé¡ãªã©ãå°ãªããã¨ï¼ããããï¼æ£ååãè¡ããã¨ã«ãã é«ãè¨è¿°è½åã¨æ±åè½åã両ç«ãã¦ãããã¨ãç¹é·ã§ããï¼ ãã¼ã¯ã¼ã: ãµãã¼ããã¯ã¿ãã·ã³ï¼æ£ååï¼ã¹ãã¼ã¹ãã¹ï¼æ°çè¨ç»æ³ï¼ æ±å Kernel machines such as the support vector machine are reviewed. Most of them are not suffered from the local optimum problem, because they
{{#tags}}- {{label}}
{{/tags}}