ã¬ã¦ã¹éç¨(Gaussian processes)ã«é¢ããã¡ã¢ï¼ï¼ï¼ æ£ç° åä¹ å¹³æ 19 å¹´ 7 æ 15 æ¥ æ¬ç¨¿ã®åèæç®ã¯ [1] ã§ãï¼ãã®ç¬¬ 2 ç« ã縮ãã¦ç´¹ä»ãã¾ããï¼ç´æ¥èªãã»ããåãããããããããã¾ããï¼ 1 éé³ãã¨ããªãç·å½¢å帰ã¢ãã« ãããã©ãã¯ããã¯ã¹ããã£ã¦ï¼p 次å ã®ãã¯ãã« x ãå ¥åã¨ãã¦ä¸ããããã¨ï¼ã²ã¨ã¤ã®æ°å¤ y ãåºåãã㨠ãã¾ãï¼ãã®ãã©ãã¯ããã¯ã¹ã®ä¸èº«ãã©ããªã£ã¦ããã®ãã説æããããã«ï¼ã¢ãã«ãã¤ãããã¨ã«ãã¾ãï¼æ¬ 稿ã§ã¯ï¼ã¬ã¦ã¹éé³ãã¨ããªãç·å½¢å帰ã¢ãã«ãèãã¾ãï¼ãã㯠f(x) = x w, y = f(x) + ε (1) ã¨ããå¼ã§æ¸ããã¨ã®ã§ããã¢ãã«ã§ãï¼ã¾ãï¼å ¥åãã¯ãã« x ã¨ï¼ã¢ãã«ã®ãã©ã¡ã¼ã¿ã§ãããã¯ãã« w ã¨ã®å ç©ãæ±ãããã¾ãï¼ããã§ã²ã¨ã¤ã®æ°å¤ f(x) ãå¾ããã¾ããï¼ããã«ã¬ã¦ã¹éé³ Îµ ã
�������� ������ ����������������� ������������������ �!#"%$&�('*) +,).-�/+10243 57698;:.<>=@?A6CB#=,DFE�G%HJI�KMLON�K>P,Q/:.? R�<>LTSVUJ6XW1B#LOYC6�<ZK�LT=@I.[ Q/:.YC6�B#S.L\K>H ]�:JRZD^<_:^=@D.<`I�P%ab:.S.LOB.U*cd6CI�e2D�:.SfP Q/:.? R�<>LTSVUJ6�PgQihkj�lfmonk[oW1B#LO=@6`Sqp�LOB.UJS^D.? [ rtstuwvxs^y;z{rt|*s*}/~�u{s{rg~_stu�~��xv � �1��C���#�f.�*�#�����*�@�J�F�g�.�F� �*�C�x�Z�{�#�*���_�,�J���.�1�`
C. E. Rasmussen & C. K. I. Williams, Gaussian Processes for Machine Learning, the MIT Press, 2006, ISBN 026218253X. c 2006 Massachusetts Institute of Technology. www.GaussianProcess.org/gpml Chapter 2 Regression Supervised learning can be divided into regression and classification problems. Whereas the outputs for classification are discrete class labels, regression is concerned with the predictio
ãªãªã¼ã¹ãé害æ å ±ãªã©ã®ãµã¼ãã¹ã®ãç¥ãã
ææ°ã®äººæ°ã¨ã³ããªã¼ã®é ä¿¡
å¦çãå®è¡ä¸ã§ã
j次ã®ããã¯ãã¼ã¯
kåã®ããã¯ãã¼ã¯
lãã¨ã§èªã
eã³ã¡ã³ãä¸è¦§ãéã
oãã¼ã¸ãéã
{{#tags}}- {{label}}
{{/tags}}