[B! ãƒ™ã‚¤ã‚º] midnightseminarã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

midnightseminar id:midnightseminar

ãƒ™ã‚¤ã‚ºã«é–¢ã™ã‚‹midnightseminarã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (22)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

è„³ã¨ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆï¼¦ï¼¡ï¼±
è„³ã®æƒ…å ±å‡¦ç†åŽŸç†ã®è§£æ˜Žã®éµã¨ãªã‚‹æŠ€è¡“ãŒ ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆã§ã™ã€‚ ã—ã‹ã—ã€å¤§åŠã®ç ”ç©¶è€…ã¯å¤§è„³çš®è³ªã¨ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆã®é®®ã‚„ã‹ãªå¯¾å¿œã«ã¤ã„ã¦ ã¾ã çŸ¥ã‚Šã¾ã›ã‚“ã€‚ è„³ã®æƒ…å ±å‡¦ç†åŽŸç†ã«åŸºã¥ã„ãŸçŸ¥èƒ½ã®é«˜ã„ãƒãƒœãƒƒãƒˆã®å®Ÿç¾ã«å‘ã‘ã€ ä¸€äººã§ã‚‚å¤šãã®ç¥žçµŒç§‘å¦è€…ãƒ»è¨ˆç®—æ©Ÿç§‘å¦è€…ã«ã€ ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆã¨å¤§è„³çš®è³ªã®é–¢ä¿‚ã‚’çŸ¥ã£ã¦ã‚‚ã‚‰ã„ãŸã„ã¨æ€ã„ã¾ã™ã€‚ ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆãã®ã‚‚ã®ã«ã¤ã„ã¦ã®è³ªå• ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆã¨ã¯ä½•ã§ã™ã‹ï¼Ÿ ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆï¼ˆãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ ã€ Bayesian networkï¼‰ ã¨ã¯ã€ç¢ºçŽ‡è«–ã«åŸºã¥ã„ãŸæŽ¨è«–ã‚’åŠ¹çŽ‡çš„ã«è¡Œã†ãŸã‚ã®æŠ€è¡“ã§ã™ã€‚ è„³ã®æ©Ÿèƒ½ã®ï¼‘ã¤ã§ã‚ã‚‹ç›´è¦³ã¨ä¼¼ãŸåƒãã‚’ã—ã¾ã™ã€‚ ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆã¯ã€è¤‡æ•°ã®äº‹è±¡ã®é–“ã®å› æžœé–¢ä¿‚ã‚’ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯æ§‹é€ ã§è¡¨ç¾ã—ã€ åŒæ™‚ã«å› æžœé–¢ä¿‚ã®å¼·ã•ã‚’è¡¨ã™æ•°å€¤ã‚‚è¨˜éŒ²ã—ãŸã‚‚ã®ã§ã™ã€‚ ã“ã®ã‚ˆã†ã«è¡¨ç¾ã•ã‚ŒãŸã€ŒçŸ¥è˜ã€ã‚’ç”¨ã„ã‚Œã°ã€å¾—ã‚‰ã‚ŒãŸè¦³æ¸¬ãƒ‡ãƒ¼ã‚¿ã«åŸºã¥ã„ã¦ æ§˜ã€…ãªäº‹è±¡ã®ç¢ºçŽ‡ã‚’ ãƒ™ã‚¤
midnightseminar 2016/11/08
ãŠã‚‚ã‚ã„

ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆ

ãƒ™ã‚¤ã‚º

ãæ©Ÿæ¢°å¦ç¿’

ãƒ‹ãƒ¥ãƒ¼ãƒ©ãƒ«ãƒãƒƒãƒˆ

â˜…
ãƒªãƒ³ã‚¯
ãƒ™ã‚¤ã‚ºéšŽå±¤è¨€èªžãƒ¢ãƒ‡ãƒ«ã«ã‚ˆã‚‹ æ•™å¸«ãªã—å½¢æ…‹ç´ è§£æž
midnightseminar 2016/02/02
ã€Œâ€œå½¢æ…‹ç´ â€ã¨ã¯ã€æ–‡å—åˆ—ã®ç”Ÿæˆç¢ºçŽ‡ã‚’æœ€å¤§ã«ã™ã‚‹ã‚ˆã†ãªæƒ…å ±ç†è«–çš„ãªå˜ä½ã¨ã—ã¦å°Žãã“ã¨ãŒã§ãã‚‹ã€

ã‘å½¢æ…‹ç´ è§£æž

NLP

ãæ©Ÿæ¢°å¦ç¿’

ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
[PDF]ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯å…¥é–€
MEDICAL IMAGING TECHNO LOGY Vol.21 No.4 September 2003 315 1 Introduction to Bayesian Network1 * Hiroki SUYARI 1. 1 1 D1, , D4 T1, , T3 Fig. 1 Fig. 1 7 8 Di Di = 1 Di = 0 Tj Tj = 1 * 263-8522 1-33 : Department of Information and Image Sciences, Faculty of Engineering, Chiba University. e-mail:[email protected] Fig. 1 One example of Bayesian Network for medical diagnosis. Tj = 0 D1, , D4, T1
midnightseminar 2015/09/14
ãƒ™ã‚¤ã‚º

ãæ©Ÿæ¢°å¦ç¿’

â˜…
ãƒªãƒ³ã‚¯
[PDF]ç¬¬10ç« ãƒžãƒ«ã‚³ãƒ•é€£éŽ–ãƒ¢ãƒ³ãƒ†ã‚«ãƒ«ãƒæ³•å…¥é–€
2019å¹´10â½‰1â½‡ çµ±è¨ˆç§‘å¦ã®ãŸã‚ã®é›»åå›³æ›¸ã‚·ã‚¹ãƒ†ãƒ ã¯ çµ±è¨ˆæ•°ç†ç ”ç©¶æ‰€ã®æ©Ÿé–¢ãƒªãƒã‚¸ãƒˆãƒªã«ç§»è¡Œã—ã¾ã—ãŸã€‚ ç§»è¡Œå¾Œã®ãƒšãƒ¼ã‚¸
midnightseminar 2015/06/21
MCMC

ãƒ™ã‚¤ã‚º

â˜…
ãƒªãƒ³ã‚¯
ãƒ¢ãƒ‡ãƒ«ã¨ã—ã¦ã®äº‹å‰åˆ†å¸ƒ - himaginaryâ€™s diary
ãƒŽã‚¢ãƒ”ãƒ‹ã‚ªãƒ³æ°ãŒçµ±è¨ˆã«ãŠã‘ã‚‹ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ã¨é »åº¦ä¸»ç¾©ã®å¯¾ç«‹ã«ã¤ã„ã¦è«–ã˜ãŸã¨ã“ã‚ã€ãƒ™ã‚¤ã‚ºçµ±è¨ˆå¦ã®æ¨©å¨ã§ã‚ã‚‹ã‚¢ãƒ³ãƒ‰ãƒªãƒ¥ãƒ¼ãƒ»ã‚²ãƒ«ãƒžãƒ³ãŒåå¿œã—ãŸã€‚ä»¥ä¸‹ã¯ãã“ã‹ã‚‰ã®æŠœç²‹ã€‚ First, a Bayesian doesnâ€™t need to stick with his or her priors, any more than any scientist needs to stick with his or her model. Itâ€™s fineâ€”indeed, recommendedâ€”to abandon or alter a model that produces implications that donâ€™t make sense (see my paper with Shalizi for a wordy discussion of this point). Second, the parall
midnightseminar 2015/05/19
ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
äº‹å‰åˆ†å¸ƒãŒãƒ™ã‚¤ã‚ºæŽ¨å®šã‚’æªã‚ã‚‹æ™‚ - himaginaryâ€™s diary
æ˜¨æ—¥ç´¹ä»‹ã—ãŸã‚²ãƒ«ãƒžãƒ³ã®ã‚¨ãƒ³ãƒˆãƒªã§ã¯ã€å¼•ç”¨éƒ¨ã®å†’é ã§äº‹å‰åˆ†å¸ƒã¸ã®å›ºåŸ·ã‚’æˆ’ã‚ã¦ã„ãŸãŒã€ãã‚Œã¯ã€ãƒŽã‚¢ãƒ”ãƒ‹ã‚ªãƒ³æ°ãŒãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ã¯ãƒ‡ãƒ¼ã‚¿ãŒä¿¡ç”¨ã§ããªã„ã¨è‡ªåˆ†ã®äº‹å‰åˆ†å¸ƒã«å›ºåŸ·ã™ã‚‹ã€ã¨æ¶æ„ã—ãŸã“ã¨ã¸ã®åè«–ã‚’æ„å›³ã—ã¦ã„ãŸã€‚ãã“ã§ãƒŽã‚¢ãƒ”ãƒ‹ã‚ªãƒ³æ°ã¯ã€äº‹å‰åˆ†å¸ƒã‚’é–“é•ãˆã‚‹ã¨ã¨ã‚“ã§ã‚‚ãªã„çµè«–ã«è‡³ã‚‹æ¥µç«¯ãªä¾‹ã¨ã—ã¦ã€çµ±è¨ˆå¦è€…ã®Cosma ShaliziãŒ4å¹´å‰ã«è‡ªãƒ–ãƒã‚°ã§ç¤ºã—ãŸã‚±ãƒ¼ã‚¹ã‚’æŒ™ã’ã¦ã„ã‚‹ï¼ˆãƒ‡ãƒãƒ³ã‚°çµŒç”±ï¼‰ã€‚ä»Šæ—¥ã¯ãã®Shaliziã®ã‚¨ãƒ³ãƒˆãƒªã‚’ç´¹ä»‹ã—ã¦ã¿ã‚‹ã€‚ The theme here is to construct some simple yet pointed examples where Bayesian inference goes wrong, though the data-generating processes are well-behaved, and the priors look harmle
midnightseminar 2015/05/19
ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
å°¤åº¦ã¨ã¯ä½•è€…ãªã®ã‹ï¼Ÿ - MyEnigma
ã¾ãšã¯ã“ã®ä¸€å†Šã‹ã‚‰ æ„å‘³ãŒã‚ã‹ã‚‹ãƒ™ã‚¤ã‚ºçµ±è¨ˆå¦ (BERET SCIENCE) ç›®æ¬¡ ç›®æ¬¡ ã¯ã˜ã‚ã« æœ€å°¤æŽ¨å®šæ³•ã¨ãƒ™ã‚¤ã‚ºæŽ¨å®šã®é•ã„ å°¤åº¦ã‚’ã‚°ãƒ©ãƒ•ã‚£ã‚«ãƒ«ã«èª¬æ˜Žã™ã‚‹è³‡æ–™ å‚è€ƒè³‡æ–™ MyEnigma Supporters ã¯ã˜ã‚ã« å¦ç”Ÿã®æ™‚ã‹ã‚‰ï¼Œ "ãã‚Œã¯æœ€å°¤æŽ¨å®šæ³•ã‚’ç”¨ã„ã¦ã„ã¾ã™" ã¨ã‹ï¼Œ "ãã®è¡Œã¯ï¼Œå°¤åº¦è¨ˆç®—ã®éƒ¨åˆ†ã§ã™ãâ€ ã¨ã‹ï¼Œã¾ã‚‹ã§å°¤åº¦ã¨ã„ã†ã‚‚ã®ã‚’çŸ¥ã£ã¦ã„ã¦ï¼Œ ä½¿ã„ã“ãªã—ã¦ã„ã‚‹ã‹ã®ã‚ˆã†ãªç™ºè¨€ã‚’ã—ã¦ãã¾ã—ãŸï¼Ž ã—ã‹ã—ï¼Œãã†è¨€ã„ãªãŒã‚‰ã‚‚ï¼Œ è‡ªåˆ†çš„ã«ã¯ï¼Œå°¤åº¦ã¨ã„ã†ã‚‚ã®ãŒã‹ãªã‚Šæ¼ ç„¶ã¨ã—ã¦ã„ã¾ã—ãŸï¼Ž ãã‚‚ãã‚‚ï¼Œå°¤åº¦ã¯æ–‡å—é€šã‚Šã€Œå°¤ã‚‚ã‚‰ã—ã•ã€ã‚’è¡¨ã™åº¦åˆã„ãªã®ã§ï¼Œ ã€Œæœ€å°¤æŽ¨å®šæ³•ã§ãƒ¢ãƒ‡ãƒ«ã®ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã‚’æ±ºå®šã—ã¾ã™ï¼Žã€ ã£ã¦ã„ã†ã®ã¯ï¼Œ ã€Œä¸€ç•ªãã‚Œã£ã½ã„ã‚‚ã®ã‚’é¸ã³ã¾ã™ã€ ã¨è¨€ã£ã¦ã„ã‚‹ã®ã¨ã‚ã¾ã‚Šå¤‰ã‚ã‚ŠãŒãªã„æ°—ãŒã—ã¦ã„ãŸã®ã§ã™(ç¬‘) æœ€è¿‘ï¼Œã“ã®æœ¬ã‚’èªã‚“ã§æœ€é©åŒ–ã‚’ä¸€ã‹ã‚‰ä½“ç³»çš„ã«å¦ã³ç›´ã—ã¦ã„ã‚‹ã®ã§ã™ãŒï¼Œ ã‚„ã£ã¨ï¼Œå°¤åº¦ã¨ã„ã†
midnightseminar 2015/04/23
ãƒ‡ãƒ¼ã‚¿ãŒé€£ç¶šé‡ã®å ´åˆã€å°¤åº¦ã¯ç¢ºçŽ‡ã§ã¯å…¨ããªã„ã¨æ€ã†ãŒã€‚

ãƒ™ã‚¤ã‚º

ã‚†å°¤åº¦

ã¨çµ±è¨ˆå¦
ãƒªãƒ³ã‚¯
MCMCã¦ã‚™ç ”ç©¶å ±å‘Š
7. è«–æ–‡æœ¬ä½“ã§å ±å‘Š äº‹ä¾‹ï¼š Regarding the clinical processâ€“mechanismâ€“outcome linkages examined in the SEM, EB results for the regression parameters in the structural portion of the SEM (see Figure 4) suggest a positive effect of the following sets of therapist interventions on change in family functioning: (a) proportion of individually focused general interventions (BE (7,9) = 17.49, 95% credible interval [6.
midnightseminar 2015/04/07
ã¨çµ±è¨ˆå¦

ãƒ™ã‚¤ã‚º

MCMC

ã‘ç ”ç©¶ãƒŽã‚¦ãƒã‚¦

â˜…

Slideshare
ãƒªãƒ³ã‚¯
æ£è¦åˆ†å¸ƒã—ãªã„ï¼ã‚«ãƒ†ã‚´ãƒªã‚«ãƒ«ãªãƒ‡ãƒ¼ã‚¿ã®æ§‹é€ æ–¹ç¨‹å¼ãƒ¢ãƒ‡ãƒ«ã§ã®æ‰±ã„æ–¹ | Theoretical Sociology
Structural Equation Modeling: A Second Course (Quantitative Methods in Education and the Behavioral Sciences) Structural Equation Modelin...ã®ä»–ã®ãƒ¬ãƒ“ãƒ¥ãƒ¼ã‚’ã¿ã‚‹Â» --- Information Age Pub Inc ï¿¥ 3,910 (2006-01-30) Sara J. Finney and Christine DiStefano, 2006, "Nonnormal and Categorical Data in Structural Equation," Gregory R. Hancock and Ralph O. Mueller (eds.) Structural Equation Modeling: A Second Course, In
midnightseminar 2015/01/25
ã„ã‚ã°ã€ãƒŽãƒ³ãƒ‘ãƒ©SEMã®è€ƒãˆæ–¹ã€‚æ½œåœ¨å¤‰æ•°ã‚’æƒ³å®šã—ã¦ãã‚ŒãŒã‚«ãƒ†ã‚´ãƒªå¤‰æ•°ã«ç¾ã‚Œã¦ã‚‹ã¨ã¿ãªã™ã¨ã„ã†ã®ã¯ã€ãªã‚‹ã»ã©ã¨ã¯æ€ã†ãŒã€æ‰±ã„ãŒé›£ã—ãã†ã â€¦ã€‚

SEM

â˜…

ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ï¼å…¥é–€ã‹ã‚‰å¿œç”¨ã¾ã§
Abstract: ( ) ( ) ( ) 1. ( ) 2. (Bayesian network, Bayesnet, belief network) [1, 2, 3, 4, 5] [6, 7, 8, 9, 10] 0 1 1 0 Xi, Xj Xi â†’ Xj Xj Xi X1 X2 X4 X3 X5 X2 0 1 X4 0 0.8 0.4 1 0.2 0.6 æ¡ä»¶ä»˜ç¢ºçŽ‡è¡¨ P(X4|X2) P(X3|X1,X2ï¼‰ ï¼°ï½ï¼ˆï¼¸ï¼“ï¼‰ ï¼°ï½ï¼ˆï¼¸ï¼•ï¼‰ ï¼°ï½ï¼ˆï¼¸ï¼”ï¼‰ P(Xï¼•|Xï¼“,Xï¼”ï¼‰ 1: Bayesian network Xj Pa(Xj) Xj Pa(Xj) ( Pa(Xj) ) P(Xj | Pa(Xj)) (1) n X1 Â· Â· Â· , Xn (2) P(X1, Â· Â· Â· , Xn) = n ï¿½ j=1 P(Xj | Pa(Xj)). (2) 1 1 3 Pa(Xj) = x1
midnightseminar 2014/11/14
ã‚ã¡ã‚ƒã‚ã¡ã‚ƒã‚ã‹ã‚Šã‚„ã™ã„ã€‚

ãæ©Ÿæ¢°å¦ç¿’

ã¨çµ±è¨ˆå¦

ãƒ™ã‚¤ã‚º

ã‚ˆäºˆæ¸¬

â˜…â˜…
ãƒªãƒ³ã‚¯
ãƒžãƒ«ã‚³ãƒ•é€£éŽ–ãƒ¢ãƒ³ãƒ†ã‚«ãƒ«ãƒæ³•ã®æœ€è¿‘ã®å±•é–‹ | CiNii Research
JaLC IRDB Crossref DataCite NDL NDL-Digital RUDA JDCat NINJAL CiNii Articles CiNii Books CiNii Dissertations DBpedia Nikkei BP KAKEN Integbio MDR PubMed LSDB Archive æ¥µåœ°ç ”ADS æ¥µåœ°ç ”å¦è¡“DB å…¬å…±ãƒ‡ãƒ¼ã‚¿ã‚«ã‚¿ãƒã‚° ãƒ ãƒ¼ãƒ³ã‚·ãƒ§ãƒƒãƒˆåž‹ç ”ç©¶é–‹ç™ºäº‹æ¥
midnightseminar 2014/09/28
ãƒ™ã‚¤ã‚º

MCMC

ã¨çµ±è¨ˆå¦
ãƒªãƒ³ã‚¯
Rã§ãƒžãƒ«ã‚³ãƒ•é€£éŽ–ãƒ¢ãƒ³ãƒ†ã‚«ãƒ«ãƒæ³•ã‚’è©¦ã™ - é¤¡åä»˜ã‚›éŒ²ã‚›
åœ°å‘³ã«ã“ã“5å¹´é–“ãã‚‰ã„ã€ãƒžãƒ«ã‚³ãƒ•é€£éŽ–ãƒ¢ãƒ³ãƒ†ã‚«ãƒ«ãƒæ³•ï¼ˆMCMCï¼‰ãŒæµè¡Œã£ã¦ã„ã‚‹ã‚ˆã†ã§ã™ã€‚æ±Žç”¨çš„ãªåˆ†å¸ƒã§ãƒ™ã‚¤ã‚ºæŽ¨å®šã‚’è¡Œã†æ™‚ã«æœ‰ç”¨ãªæ•°å€¤è§£æžã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®ç·ç§°ã§ã€Metropolis-Hastings algorithmï¼ˆM-Hã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ï¼‰ãªã©ãŒä¸»è¦ãªãƒ¡ã‚½ãƒƒãƒ‰ã¨ã—ã¦ä½¿ã‚ã‚Œã¦ã„ã¾ã™ã€‚ãŸã ã—ã€ãƒ™ã‚¤ã‚ºæŽ¨å®šä»¥å¤–ã§ã‚‚åˆ©ç”¨ã™ã‚‹äº‹ã¯ã§ãã¾ã™ã€‚ Rã§ã‚‚MCMCpackã¨è¨€ã†ãƒ‘ãƒƒã‚±ãƒ¼ã‚¸ãŒã‚ã‚‹ã®ã§ã™ãŒã€å–æ‰±èª¬æ˜Žæ›¸ã‚’è¦‹ã‚‹é™ã‚Šã€ãƒ™ã‚¤ã‚ºæŽ¨å®šãŒå‰æã¨ãªã£ã¦ãŠã‚ŠM-Hã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã ã‘ã‚’è©¦ã™ã“ã¨ã¯é›£ã—ãã†ã§ã™ï¼ˆè¿½è¨˜ï¼šMCMCmetrop1R()é–¢æ•°ã§åˆ©ç”¨ã§ãã¾ã™ï¼‰ã€‚ã‚‚ã£ã¨ã‚‚åŒã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã¯æ¯”è¼ƒçš„ã‚·ãƒ³ãƒ—ãƒ«ãªãƒ¢ãƒŽãªã®ã§ã€ãƒã‚¢ã‚½ãƒ³åˆ†å¸ƒã‹ã‚‰ä¹±æ•°ã‚’ä½œæˆã—ã€ãã‚Œã‚’æŽ¨å®šã™ã‚‹ç·´ç¿’ã‚’ã—ã¦ã¿ã¾ã—ãŸã€‚ 1. ãƒã‚¢ã‚½ãƒ³åˆ†å¸ƒã‹ã‚‰Î»=1ã®ä¹±æ•°ã‚’ä½œæˆ ãƒã‚¢ã‚½ãƒ³åˆ†å¸ƒã‹ã‚‰Î»=1ã®ä¹±æ•°ã‚’ä½œæˆã—ã¾ã™ã€‚ç·´ç¿’ã®ãŸã‚ã«ãƒã‚¢ã‚½ãƒ³åˆ†å¸ƒã‚’åˆ©ç”¨ã—ãŸã®ã¯ã€æŽ¨å®šã™ã‚‹ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ãƒ¼ãŒ
midnightseminar 2014/09/28
MCMC

Rè¨€èªž

ãƒ™ã‚¤ã‚º

uncorrelated
ãƒªãƒ³ã‚¯
ãƒ™ã‚¤ã‚ºæŽ¨å®šã‚’çŸ¥ã£ã¦ã„ã‚‹ãƒ•ãƒªã‚’ã™ã‚‹ãŸã‚ã®çŸ¥è˜
æœ€è¿‘ã¯ãƒ™ã‚¤ã‚¸ã‚¢ãƒ³ãŒå¢—ãˆã¦ãã¦ã€å®Ÿç”¨åˆ†é‡Žã§ã®åˆ©ç”¨ã‚‚é€²ã‚“ã§ã„ã‚‹ã‚ˆã†ã ã€‚è©±é¡Œã¨ã—ã¦ã¯çŸ¥ã£ã¦ãŠããŸã„ãŒã€ä¸–é–“ä¸€èˆ¬ã«ã¯ç†è§£ã«æ··ä¹±ã‚’ç”Ÿã‚“ã§ã„ã‚‹ã‚ˆã†ã ã€‚ ãƒ™ã‚¤ã‚ºæŽ¨å®šã¯å…¥é–€ãƒ¬ãƒ™ãƒ«ã®çµ±è¨ˆå¦ã®æ•™ç§‘æ›¸ã§ã¯ã‚ªãƒžã‚±çš„ãªæ‰±ã„ãŒã•ã‚Œã¦ãŠã‚Šã€å®Ÿéš›ã«ä¼çµ±çš„ãªçµ±è¨ˆæ‰‹æ³•ã‚’æ‹¡å¼µã—ã¦ã„ã‚‹é¢ãŒå¼·ã„ã€‚ãã†ã„ã†æ„å‘³ã§ã¯ã€èª¤è§£ã‚„æ··ä¹±ãŒã‚ã£ã¦ã‚‚ä»•æ–¹ãŒç„¡ã„ã€‚ åˆ©ç”¨ã™ã‚‹å¿…è¦ãŒã‚ã‚‹ã®ã‹ç„¡ã„ã®ã‹è‰¯ãåˆ†ã‹ã‚‰ãªã„ç‚¹ã‚‚å¤šã„ã®ã ãŒã€çŸ¥ã‚‰ãªã„ã¨å‘Šç™½ã™ã‚‹ã®ã‚‚æ°—æ¥ãšã‹ã—ã„ã‹ã‚‚çŸ¥ã‚Œãªã„ã€‚è‡ªåˆ†ã§ã¯ãƒ™ã‚¤ã‚ºæŽ¨å®šã§åˆ†æžã‚’è¡Œã‚ãªã„äººãŒã€ãƒ™ã‚¤ã‚ºä¿¡è€…ã¨è©±ã‚’ã‚ã‚ã›ã‚‹ãŸã‚ã«æœ€ä½Žé™çŸ¥ã£ã¦ãŠãã¹ãäº‹ã‚’ã¾ã¨ã‚ã¦ã¿ãŸã€‚ 1. ãƒ™ã‚¤ã‚ºæŽ¨å®šã¨ã¯ä½•ã‹ï¼Ÿ ãƒ™ã‚¤ã‚ºæŽ¨å®šã¨ã¯ã€ãƒ™ã‚¤ã‚ºã®å®šç†ã‚’å¿œç”¨ã—ãŸæŽ¨å®šæ‰‹æ³•ã ã€‚ç«¯çš„ã«ç†è§£ã™ã‚‹ãŸã‚ã«ã¯ã€æœ€å°¤æ³•ã«äº‹å‰ç¢ºçŽ‡ã‚’å°Žå…¥ã—ã¦ã„ã‚‹äº‹ã ã‘è¦šãˆã‚Œã°è‰¯ã„ã€‚ã“ã‚Œã§å“²å¦çš„è°è«–ã‚’å…¨ã¦å›žé¿ã—ã¦ãƒ™ã‚¤ã‚ºæŽ¨å®šã‚’æŠŠæ¡ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ã€‚ ä¸‹ã®(1)å¼ã§ã¯Ï€ï¼ˆÎ¸ï¼‰ãŒäº‹å‰ç¢ºçŽ‡ã€Ï€(Î¸|x)ãŒäº‹å¾Œç¢º
midnightseminar 2014/09/28
â€œæœ€å°¤æ³•ã®æ‹¡å¼µã ã¨ç†è§£ã—ã¦ã—ã¾ã†ã¨èº«è¿‘ã«ãªã‚‹ã€‚åˆ©ç”¨ç›®çš„ã‚‚ä¸€èˆ¬ã«ã¯æŽ¨å®šé‡ã®ä¿¡é ¼æ€§ã‚’é«˜ã‚ã‚‹ãŸã‚ã ã¨æ€ã£ã¦ãŠã‘ã°è‰¯ã„â€

uncorrelated

â˜…â˜…

ã¨çµ±è¨ˆå¦

ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
ã‚·ãƒ£ãƒãƒ³ãƒ»ãƒãƒ¼ãƒãƒ¥ãƒ»ãƒžã‚°ãƒ¬ã‚¤ãƒ³ã€Žç•°ç«¯ã®çµ±è¨ˆå¦ãƒ™ã‚¤ã‚ºã€ - logical cypher scape2
ãƒ™ã‚¤ã‚ºæ´¾çµ±è¨ˆå¦ã®æ´å²ã«ã¤ã„ã¦ã®æœ¬ ã‚ãã¾ã§æ´å²ãªã®ã§ã€ãƒ™ã‚¤ã‚ºçµ±è¨ˆã«ã¤ã„ã¦ã®èª¬æ˜Žã¯ã‚ã¾ã‚Šãªã„ãŒã€é–¢ã‚ã£ãŸäººç‰©ã®ã‚¨ãƒ”ã‚½ãƒ¼ãƒ‰ãªã©ãŒä¸å¿ƒã¨ãªã£ã¦ã„ã‚‹ã€‚ ã‚¢ãƒ¬ãƒƒã‚¯ã‚¹ãƒ»ãƒãƒ¼ã‚¼ãƒ³ãƒãƒ¼ã‚°ã€Žç§‘å¦å“²å¦ã€ - logical cypher scape2ã‚„ãƒ©ãƒ—ãƒ©ã‚¹ã€Žç¢ºçŽ‡ã®å“²å¦çš„è©¦è«–ã€ï¼ˆå†…äº•æƒ£ä¸ƒè¨³ï¼‰ - logical cypher scape2ãŒã€ãƒ™ã‚¤ã‚ºã ã£ãŸã®ã§ã€ãƒ™ã‚¤ã‚ºã®å‹‰å¼·ã€‚ çµ±è¨ˆå¦ã®ä¸ã§ã¯ã€é »åº¦ä¸»ç¾©ã¨ã„ã†ã®ãŒä¸»æµã§ã€ãƒ™ã‚¤ã‚ºæ´¾ã¯ãšã£ã¨ãƒžã‚¤ãƒŽãƒªãƒ†ã‚£ã ã£ãŸã€‚ ã¾ãšã€ãƒ™ã‚¤ã‚ºè‡ªèº«ãŒãã‚‚ãã‚‚ãƒ™ã‚¤ã‚ºã®å®šç†ã‚’å…¬è¡¨ã—ã¦ãŠã‚‰ãšã€ç¶šã„ã¦æ•°å¦çš„ãªå®šå¼åŒ–ã‚’è¡Œã£ãŸãƒ©ãƒ—ãƒ©ã‚¹ã‚‚æ™©å¹´ã¯é »åº¦ä¸»ç¾©ã«è»¢å‘ã—ã€19ä¸–ç´€ã¯å®Œå…¨ã«é »åº¦ä¸»ç¾©ã®æ™‚ä»£ã§ã€ç¬¬äºŒæ¬¡å¤§æˆ¦ã§å®Ÿéš›ã«ä½¿ã‚ã‚Œã‚‹ã‚ˆã†ã«ãªã‚‹ã‚‚è»äº‹æ©Ÿå¯†æ‰±ã„ã§ä¸–ã®ä¸ã«åºƒã¾ã‚‰ãšã€æˆ¦å¾Œå°‘ã—ãšã¤ãƒ™ã‚¤ã‚ºæ´¾ãŒåºƒã¾ã‚Šå§‹ã‚ã‚‹ã‚‚è¨ˆç®—ãŒå¤§å¤‰ã§ãªã‹ãªã‹å®Ÿç”¨åŒ–ã•ã‚Œãšâ€¦â€¦ ã¨ã€ãƒ™ã‚¤ã‚ºçš„ãªè€ƒãˆæ–¹ã¯ãšã£ã¨ä¸é‡ã®æ—¥ã€…ã‚’é€ã£ã¦ããŸã‚ã‘ã ãŒã€ãã®
midnightseminar 2014/07/28
ãƒ™ã‚¤ã‚º

â˜…
ãƒªãƒ³ã‚¯
ã€€ã€Œç•°ç«¯ã®çµ±è¨ˆå¦ãƒ™ã‚¤ã‚ºã€ - shorebirdã€€é€²åŒ–å¿ƒç†å¦ä¸å¿ƒã®æ›¸è©•ãªã©
ç•°ç«¯ã®çµ±è¨ˆå¦ ãƒ™ã‚¤ã‚º ä½œè€…: ã‚·ãƒ£ãƒãƒ³ãƒ»ãƒãƒ¼ãƒãƒ¥ãƒžã‚°ãƒ¬ã‚¤ãƒ³,Sharon Bertsch McGrayne,å†¨æ°¸æ˜Ÿå‡ºç‰ˆç¤¾/ãƒ¡ãƒ¼ã‚«ãƒ¼: è‰æ€ç¤¾ç™ºå£²æ—¥: 2013/10/23ãƒ¡ãƒ‡ã‚£ã‚¢: å˜è¡Œæœ¬ã“ã®å•†å“ã‚’å«ã‚€ãƒ–ãƒã‚° (28ä»¶) ã‚’è¦‹ã‚‹ æœ¬æ›¸ã¯ãƒ™ã‚¤ã‚ºçµ±è¨ˆå¦ã®å¦èª¬å²ã«ã‹ã‹ã‚‹æœ¬ã§ï¼Œã‚¢ãƒ¡ãƒªã‚«ã®ã‚µã‚¤ã‚¨ãƒ³ã‚¹ãƒ©ã‚¤ã‚¿ãƒ¼ã®æ‰‹ã«ã‚ˆã‚‹ã‚‚ã®ï¼ŽåŽŸé¡Œã¯ã€ŒThe Theory That Would Not Die: How Beyesâ€™ Rule Cracked the Enigma Code, Hunted Down Russian Submarines, and Emerged Triumphant from Two Centuries of Controversyã€ã¨ã„ã†ã“ã¨ã§ã¾ã‚‹ã§ã‚¹ãƒ‘ã‚¤å°èª¬ã®ã‚ˆã†ãªã‚¿ã‚¤ãƒˆãƒ«ã ãŒï¼Œé‚¦é¡Œã¯ã‚ˆã‚Šè«–äº‰å²ã‚’æ„è˜ã—ãŸã‚‚ã®ã«ãªã£ã¦ã„ã‚‹ï¼Žç§ã¨ã—ã¦ã¯ã‚½ãƒ¼ãƒãƒ¼ã®ã€Œç§‘å¦ã¨è¨¼æ‹ ã€ã‚’èªã‚“ã§ãƒ™ã‚¤ã‚ºä¸»ç¾©ã¨é »åº¦ä¸»ç¾©
midnightseminar 2014/07/18
ã€Œç‰äº‹å‰ç¢ºçŽ‡ã®ä¸»è¦³æ€§ã®å•é¡ŒãŒå¿…è¦ä»¥ä¸Šã«ãƒ•ã‚©ãƒ¼ã‚«ã‚¹ã•ã‚ŒãŸã€ã€Œãƒ•ã‚£ãƒƒã‚·ãƒ£ãƒ¼ï¼Œãƒã‚¤ãƒžãƒ³ãŒè‡ªèª¬ã«å¯¾ç«‹ã™ã‚‹ã‚‚ã®ã«åç‹ã ã£ãŸã€ã€Œãƒ™ã‚¤ã‚ºã®è¼ã‹ã—ã„æˆåŠŸä¾‹ãŒè»äº‹æ©Ÿå¯†æŒ‡å®šã«ã‚ˆã‚Šä¸–é–“ã«çŸ¥ã‚‰ã‚Œã‚‹ã“ã¨ãŒãªã‹ã£ãŸã€

ã¨çµ±è¨ˆ

ãŒå¦èª¬å²

â˜…â˜…

æ›¸è©•

ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
Stanã§çµ±è¨ˆãƒ¢ãƒ‡ãƒªãƒ³ã‚°ã‚’å¦ã¶(5): ã¨ã‚Šã‚ãˆãšéšŽå±¤ãƒ™ã‚¤ã‚ºãƒ¢ãƒ‡ãƒ«ã‚’è©¦ã—ã¦ã¿ã‚‹ï¼ˆå¿œç”¨ç·¨ï¼šãƒˆãƒ¬ãƒ³ãƒ‰ã®ã‚ã‚‹ãƒ¢ãƒ‡ãƒ«ï¼‰ *è¿½è¨˜2ä»¶ã‚ã‚Š - æ¸‹è°·é§…å‰ã§åƒããƒ‡ãƒ¼ã‚¿ã‚µã‚¤ã‚¨ãƒ³ãƒ†ã‚£ã‚¹ãƒˆã®ãƒ–ãƒã‚°
ã“ã®ã‚·ãƒªãƒ¼ã‚ºè¨˜äº‹ã€å…¨ç„¶çœŸé¢ç›®ã«äº‹å‰åˆ†å¸ƒã®å‹‰å¼·ã‚’ã—ã¦ã„ãªã„äººé–“ãŒStanã§ç„¡ç†ã‚„ã‚Šãƒ•ãƒ«ãƒ™ã‚¤ã‚ºã‚’ã‚„ã‚ã†ã¨ã„ã†ç„¡è¬€ãªä»£ç‰©ã§ã”ã–ã„ã¾ã™ãŒã€ä½•ã ã‹ã‚“ã ã§æ®µã€…ä½³å¢ƒã«å…¥ã£ã¦ã¾ã„ã‚Šã¾ã—ãŸã€‚ ã¨ã„ã†ã“ã¨ã§ã€ä»Šå›žã¯éšŽå±¤ãƒ™ã‚¤ã‚ºãƒ¢ãƒ‡ãƒ«ã‚’ã“ã‚“ãªæ„Ÿã˜ã§ã‚„ã£ã¦ã¿ã¾ã—ãŸã¨ã„ã†ä¾‹ã‚’æŒ™ã’ã¦ã¿ã‚ˆã†ã‹ã¨æ€ã„ã¾ã™ã€‚ã¡ãªã¿ã«å†…å®¹çš„ã«ã¯@berobero11ã•ã‚“ã®ã“ã¡ã‚‰ã®è¨˜äº‹ï¼ˆã€ŒRStanã§ã€Žäºˆæ¸¬ã«ã„ã‹ã™çµ±è¨ˆãƒ¢ãƒ‡ãƒªãƒ³ã‚°ã®åŸºæœ¬ã€ã®å£²ä¸Šãƒ‡ãƒ¼ã‚¿ã®åˆ†æžã‚’ãƒˆãƒ¬ãƒ¼ã‚¹ã—ã¦ã¿ãŸã€ï¼‰ã‚’ã‚°ãƒ¬ãƒ¼ãƒ‰ãƒ€ã‚¦ãƒ³ã•ã›ãŸæ„Ÿã˜ã®ã‚‚ã®ã ã£ãŸã‚Šã—ã¾ã™ï¼ˆç¬‘ï¼‰ã€‚ãã—ã¦å…ˆæ—¥æ‹›å¾…è¬›æ¼”ã•ã›ã¦ã„ãŸã ã„ãŸæ™‚ã®æœ€å¾Œã®æ–¹ã§å–ã‚Šä¸Šã’ãŸä¾‹ã§ã‚‚ã‚ã‚Šã¾ã™ã€‚ ãã‚“ãªã‚ã‘ã§ã€ã©ã®ã‚ˆã†ã«ã—ã¦ã‚„ã£ã¦ã„ã£ãŸã‹ã‚’å«ã‚ã¦ã‚µã‚¯ã‚µã‚¯è¦‹ã¦ã„ãã¾ã—ã‚‡ã†ã€‚éšŽå±¤ãƒ™ã‚¤ã‚ºã«ã¤ã„ã¦å¿˜ã‚Œã¡ã‚ƒã£ãŸã¨ã„ã†äººã¯ã€å‰å›žã®è¨˜äº‹ã‚ãŸã‚Šã‚’èªã‚“ã§å¾©ç¿’ã—ã¦ã‚‚ã‚‰ãˆã‚Œã°ã€‚ ãƒ‡ãƒ¼ã‚¿ã‚’ã‚¤ãƒ³ãƒãƒ¼ãƒˆã™ã‚‹ ã„ã¤ã‚‚é€šã‚Šã€ã‚µãƒ³ãƒ—ãƒ«ãƒ‡ãƒ¼ã‚¿ã‚’GitHubã«ä¸Šã’ã¦ã‚ã‚‹ã®ã§æŒ
midnightseminar 2014/06/12
ãƒ™ã‚¤ã‚º

Rè¨€èªž

ã˜æ™‚ç³»åˆ—åˆ†æž
ãƒªãƒ³ã‚¯
[PDF]å§‹ã‚ã‚ˆã†ï¼ãƒ™ã‚¤ã‚ºæŽ¨å®šã«ã‚ˆã‚‹ ãƒ‡ãƒ¼ã‚¿è§£æž
midnightseminar 2014/05/20
ãƒ™ã‚¤ã‚ºæŽ¨å®šã®èª¬æ˜Žã¨ã—ã¦ã‚ã¡ã‚ƒã‚ã¡ã‚ƒåˆ†ã‹ã‚Šã‚„ã™ã„ã€‚

ãƒ™ã‚¤ã‚º

â˜…
ãƒªãƒ³ã‚¯
ã€Œãƒ‡ãƒ¼ã‚¿è§£æžã®ãŸã‚ã®çµ±è¨ˆãƒ¢ãƒ‡ãƒªãƒ³ã‚°å…¥é–€ã€èªæ›¸ãƒŽãƒ¼ãƒˆ - Wolfeyes Bioinformatics beta
ã€Œãƒ‡ãƒ¼ã‚¿è§£æžã®ãŸã‚ã®çµ±è¨ˆãƒ¢ãƒ‡ãƒªãƒ³ã‚°å…¥é–€ã€ã‚’ã²ã¨é€šã‚Šèªã‚“ã ï¼Žæœ¬æ›¸ã¯GLMã‹ã‚‰MCMCã«ã‚ˆã‚‹åˆ†å¸ƒæŽ¨å®šã¾ã§ã®ä¸€é€£ã®çµ±è¨ˆãƒ¢ãƒ‡ãƒªãƒ³ã‚°ã®æµã‚Œã‚’ï¼Œç”Ÿæ…‹å¦ã«ãŠã‘ã‚‹ç ”ç©¶ã®å•é¡Œã«å³ã—ãŸãƒ†ã‚¹ãƒˆãƒ‡ãƒ¼ã‚¿ã‚„Rã‚’ä½¿ã£ãŸè§£æžä¾‹ã¨ã¨ã‚‚ã«è§£èª¬ã—ãŸæœ¬ã§ã‚ã‚‹ï¼Žæœ¬æ›¸ã‚’æ›¸ã‹ã‚ŒãŸä¹…ä¿ã•ã‚“ã®è¬›ç¾©è³‡æ–™ã¯å‰ã€…ã‹ã‚‰æ‹è¦‹ã—ã¦ã„ãŸã®ã ãŒï¼Œä»Šå›žã¯ãã‚ŒãŒå…¨ä½“ã‚’é€šã—ã¦éžå¸¸ã«ã‚ˆãã¾ã¨ã¾ã£ã¦ã„ã‚‹å°è±¡ã‚’å—ã‘ãŸï¼Žã‚„ã¯ã‚Šå®Ÿä¾‹ã«æ²¿ã£ãŸä¾‹é¡ŒãŒã‚ã£ã¦ï¼Œãã‚Œã‚’è§£æ±ºã™ã‚‹ãŸã‚ã®ã‚¹ãƒˆãƒ¼ãƒªãƒ¼ãŒçµ„ã¾ã‚Œã¦ã„ã‚‹ã¨ï¼Œä½•ãŒå•é¡Œã§ä½•ã‚’ã™ã¹ããªã®ã‹ï¼Œãã—ã¦ãã®è©•ä¾¡æ–¹æ³•ã‚’å«ã‚ã¦ãƒãƒƒã‚ãƒªã¨ã—ã¦ã„ã¦èªã¿ã‚„ã™ã„ï¼ŽRã®ã‚³ãƒ¼ãƒ‰ã«é–¢ã—ã¦ã‚‚ï¼Œã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ–¹æ³•ã‹ã‚‰ãã®è§£é‡ˆã®ä»•æ–¹ã¾ã§ä¸å¯§ã«è§£èª¬ãŒçµ„ã¾ã‚Œã¦ãŠã‚Šï¼Œæ•°å¼ã¨ã®ã¤ãªãŒã‚Šã‚‚ã‚ã‹ã‚Šã‚„ã™ã„ï¼Žå€‹äººçš„ã«ã¯ï¼Œå¾ŒåŠã®MCMCã®å®Ÿé¨“ã«é–¢ã—ã¦ã¯å¤§éƒ¨åˆ†ã‚’WinBUGSã«ãŠä»»ã›ã§ï¼Œã‚·ãƒŸãƒ¥ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ã®éŽç¨‹ãŒå°‘ã—ä¸æ˜Žçžã ã£ãŸæ„Ÿã˜ã‚‚ã™ã‚‹ã®ã ãŒï¼Œé™ã‚‰ã‚ŒãŸç´™é¢ã§MCMCã®ç´°ã‹ã„å®Ÿè£…æ–¹æ³•
midnightseminar 2014/05/17
ç·‘æœ¬ã®å¦ç¿’ãƒŽãƒ¼ãƒˆã€‚

ã¿ç·‘æœ¬

â˜…

ã¨çµ±è¨ˆ

GLM

MCMC

ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
ãƒ™ã‚¤ã‚ºçµ±è¨ˆå…¥é–€
7. è£œè¶³ ãƒ‡ãƒ¼ã‚¿è§£æžã¨ãƒ¢ãƒ‡ãƒ« ãƒ‡ãƒ¼ã‚¿ã®è§£æžï¼çµ±è¨ˆãƒ¢ãƒ‡ãƒªãƒ³ã‚° çµ±è¨ˆãƒ¢ãƒ‡ãƒ«ã¨ã¯è¦³æ¸¬ãƒ‡ãƒ¼ã‚¿ã®ãƒ‘ã‚¿ãƒ¼ãƒ³ã‚’ã†ã¾ãèª¬æ˜Žã§ãã‚‹ãƒ¢ãƒ‡ãƒ« åŸºæœ¬ã¯ç¢ºçŽ‡åˆ†å¸ƒã§ã‚ã‚Šã€ãã®ç¢ºçŽ‡åˆ†å¸ƒã®å½¢ã‚’æ±ºå®šã™ã‚‹ã‚‚ã®ãŒãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ ã€ŒæŽ¨å®šã€ï¼ãƒ¢ãƒ‡ãƒ«ã®å½“ã¦ã¯ã¾ã‚ŠãŒè‰¯ããªã‚‹ã‚ˆã†ã«ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã‚’æ±ºå®šã™ã‚‹ã“ã¨ å›žå¸°ç³»ãƒ¢ãƒ‡ãƒ« GLMï¼šæŒ‡æ•°é–¢æ•°æ— ãƒãƒ¯ã‚½ãƒ³å›žå¸° ãƒã‚¸ã‚¹ãƒ†ã‚£ãƒƒã‚¯å›žå¸° GLMM(Generalized linear mixed model) â€¦éŽåˆ†æ•£ãƒ‡ãƒ¼ã‚¿ã®åˆ†æžãªã© éšŽå±¤ãƒ™ã‚¤ã‚º 7 8. ãƒ™ã‚¤ã‚ºã®å®šç† 1 ãƒ™ã‚¤ã‚ºã®å®šç†ã¨ã¯ã€ p y|H p(H) p(H|y) = p(y) â‡” äº‹å¾Œæƒ…å ± = ãƒ‡ãƒ¼ã‚¿ã®æƒ…å ± + äº‹å‰æƒ…å ± yï¼šä½•ã‚‰ã‹ã®æƒ…å ± Hï¼šä½•ã‚‰ã‹ã®ä»®èª¬ p(H) äº‹å‰ã®ç¢ºä¿¡åº¦ ä»®èª¬HãŒæ£ã—ã„ã¨ã„ã†äº‹å‰ã®ç¢ºä¿¡åº¦ p(y|H) å°¤åº¦é–¢æ•° HãŒçœŸã®éš›ã«ãƒ‡ãƒ¼ã‚¿ãŒå¾—ã‚‰ã‚Œã‚‹å°¤åº¦ p(H|y) äº‹å¾Œã®ç¢ºä¿¡åº¦ æƒ…å ±yãŒå¾—ã‚‰ã‚ŒãŸå¾Œã®ä»®èª¬Hã®ç¢ºä¿¡åº¦ P
midnightseminar 2014/05/10
ã¨çµ±è¨ˆ

ãƒ™ã‚¤ã‚º
ãƒªãƒ³ã‚¯
å¾“æ¥ã®æŽ¨å®šæ³•ã¨ãƒ™ã‚¤ã‚ºæŽ¨å®šæ³•ã®é•ã„ | Sunny side up!
ãƒ™ã‚¤ã‚ºæŽ¨å®šã£ã¦ã€æœ€è¿‘ã¯ã‚„ã£ã¦ãã¦ã¾ã™ãã€‚åƒ•ã‚‚æµè¡Œã‚Šã«ãŠã„ã¦è¡Œã‹ã‚Œãªã„ã‚ˆã†ã«å‹‰å¼·ã—ã¦ã„ã¾ã™ã€‚ ç†è«–çš„ãªè©±ã‚„æ•°å¦çš„ãªè©±ã¯ã„ã‚ã„ã‚Webã‚„æœ¬ã‚’ã‚ã•ã‚Œã°å‡ºã¦ãã¾ã™ãŒã€å®Ÿç”¨é¢ã¨ã‹è§£é‡ˆé¢ã«ã¤ã„ã¦è¨€åŠã—ã¦ã„ã‚‹ã‚‚ã®ã¯å°‘ãªã„ã§ã™ãã€‚ ä»Šå›žã¯æ¸…æ°´ã®å€‹äººçš„ãªæ„è¦‹ã¨ã—ã¦ã€ãƒ™ã‚¤ã‚ºãŒã©ã†ã„ã†é¢¨ã«ä½¿ãˆãã†ã‹æ›¸ã„ã¦ã¿ã¾ã™ã€‚æ•°å¦çš„ãªè©±ã¯ãªã—ã§ã€‚ã‚ˆãã‚ã‹ã‚‰ãªã„ã®ã§ã€‚ èˆˆå‘³ã‚ã‚‹äººã¯ç¶šãã‚’ã©ã†ãžã€‚ 2016/2/1è¿½è¨˜ï¼šãƒ™ã‚¤ã‚ºçµ±è¨ˆã«ã¤ã„ã¦ï¼Œå…¥é–€çš„ãªè³‡æ–™ã‚’ä½œã‚Šã¾ã—ãŸã€‚å¿ƒç†å¦è€…ã®ãŸã‚ã®ãƒ™ã‚¤ã‚ºçµ±è¨ˆå…¥é–€ã‚‚ã‚ã‚ã›ã¦ã©ã†ãžã€‚ ãƒ™ã‚¤ã‚ºæŽ¨å®šæ³•ã®å‰ã«ã€å¾“æ¥æ³•ã®ä»£è¡¨ã¨ã—ã¦æœ€å°¤æŽ¨å®šæ³•ã«ã¤ã„ã¦è§¦ã‚Œã¦ãŠãã¾ã™ã€‚ ãã®æ–¹æ³•ã¨ãƒ™ã‚¤ã‚ºãŒã©ã†é•ã†ã®ã‹ã«ã¤ã„ã¦ã€ãã®ã‚ã¨è¿°ã¹ã¾ã™ã€‚ æœ€å°¤æŽ¨å®šæ³• æœ€å°¤æ³•ã¨ã‚‚ã„ã‚ã‚Œã¾ã™ãŒã€åŸºæœ¬çš„ãªç™ºæƒ³ã¯ã€ãƒ¢ãƒ‡ãƒ«ã¨ãƒ‡ãƒ¼ã‚¿ã®é–¢ä¿‚ã‚’æ¬¡ã®ã‚ˆã†ã«è€ƒãˆã¾ã™ã€‚ çœŸã®ãƒ¢ãƒ‡ãƒ«ã¨ã„ã†ã®ãŒã‚ã£ã¦ã€æˆ‘ã€…ã¯ãã®ãƒ¢ãƒ‡ãƒ«ã‹ã‚‰ç™ºç”Ÿã—ãŸãƒ‡ãƒ¼ã‚¿ã‚’æ‰‹ã«å…¥ã‚Œã¦ã„ã‚‹ã€‚çœŸå€¤ã¯ä¸€ã¤
midnightseminar 2014/05/01
ã¨çµ±è¨ˆ

ãƒ™ã‚¤ã‚º

â˜…
ãƒªãƒ³ã‚¯
1 2 æ¬¡ã®ãƒšãƒ¼ã‚¸