æ¬å½ã¯ä¸ãæ¥ä¸ã«ã¾ã¨ããªã¨ã³ããªã1æ¬ãããã¯æ¸ãäºå®ã ã£ãã®ã ããã¡ãã£ã¨ç¡çã ã£ããã¨ããããã§ãå®è³ªçã«æ°å¹´ä¸æ¬ç®ã®ã¨ã³ããªã§ããLarge Scale Learning to Rank (D. Sculley, NIPS Workshop on Advances in Ranking, 2009) (pdf) ãèªãã ã®ã§ã1æ¬ç®ã®ã¨ã³ããªã¨ãã¦ãã®è«æãç´¹ä»ãããã ã§ã¯æ©éæ¬é¡ã«å ¥ãããé ä½å¦ç¿ã«ããã¦ãPairwise Learningãåç´ã«è¡ãã¨ãn^2ã®å¦ç¿ã³ã¹ããããããããã¯è¨ç®æéã¨ãã¦ã¯å³ããé¨é¡ã«å ¥ããããããé ä½å¦ç¿ã£ã¦ãªã«ãã¨ãã人ã¯ãWWW2009ã®ãã¥ã¼ããªã¢ã«(pdf)ã¨ããåç §ãã¦ãã ããã Bottouãã¯ãSGDã®ä¸è¬åè½åã¯ãã¼ã¿ã»ããã®ãµã¤ãºã«ä¾ãããã©ãã ãã®stochastic stepãå®è¡ãããã§æ±ºã¾ãã¨è¨ãäºã示ãããããã§ãSc
{{#tags}}- {{label}}
{{/tags}}