東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Applications of saddlepoint approxim...

Yang, Bo.

FindBook

Google Book

Amazon

博客來

Applications of saddlepoint approximations for small sample inference.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Applications of saddlepoint approximations for small sample inference./
作者:	Yang, Bo.
面頁冊數:	95 p.
附註:	Source: Dissertation Abstracts International, Volume: 61-01, Section: B, page: 0035.
Contained By:	Dissertation Abstracts International61-01B.
標題:	Biology, Biostatistics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9959912
ISBN:	0599632240

Applications of saddlepoint approximations for small sample inference.
Yang, Bo.

Applications of saddlepoint approximations for small sample inference. - 95 p.

Source: Dissertation Abstracts International, Volume: 61-01, Section: B, page: 0035.

Thesis (Ph.D.)--The University of Rochester, 2000.

There has been remarkable progress in asymptotic methods for inference in the past decade. Recently developed asymptotics based on saddlepoint methods provide important practical methods for inference in multiparameter exponential families, and especially in generalized linear models. Introduced by Daniels in 1954, the saddlepoint method was first used to improve the central limit theorem and direct Edgeworth expansions. Barndorff-Nielsen and Cox (1979) revived interest in this, in terms of parametric statistical models, with some developments for conditional distributions eliminating nuisance parameters. More recently many statistical applications of the saddlepoint approximation have been developed. These include the approximations of Barndorff-Nielsen (1983) to the distribution of the maximum likelihood estimate, Lugannani and Rice (1980) to the univariate tail probability, Skovgaard (1987) to the conditional tail probability, and the extension of Davison (1988) to generalized linear models.

ISBN: 0599632240Subjects--Topical Terms:

1018416
Biology, Biostatistics.

Applications of saddlepoint approximations for small sample inference.
LDR:02852nam 2200289 a 45 001 929285
005 20110427
008 110427s2000 eng d
020 $a 0599632240
035 $a (UnM)AAI9959912
035 $a AAI9959912
040 $a UnM $c UnM
100 1 $a Yang, Bo. $3 1252769
245 1 0 $a Applications of saddlepoint approximations for small sample inference.
300 $a 95 p.
500 $a Source: Dissertation Abstracts International, Volume: 61-01, Section: B, page: 0035.
500 $a Supervisor: John E. Kolassa.
502 $a Thesis (Ph.D.)--The University of Rochester, 2000.
520 $a There has been remarkable progress in asymptotic methods for inference in the past decade. Recently developed asymptotics based on saddlepoint methods provide important practical methods for inference in multiparameter exponential families, and especially in generalized linear models. Introduced by Daniels in 1954, the saddlepoint method was first used to improve the central limit theorem and direct Edgeworth expansions. Barndorff-Nielsen and Cox (1979) revived interest in this, in terms of parametric statistical models, with some developments for conditional distributions eliminating nuisance parameters. More recently many statistical applications of the saddlepoint approximation have been developed. These include the approximations of Barndorff-Nielsen (1983) to the distribution of the maximum likelihood estimate, Lugannani and Rice (1980) to the univariate tail probability, Skovgaard (1987) to the conditional tail probability, and the extension of Davison (1988) to generalized linear models.
520 $a This dissertation will explore applications of saddlepoint approximations for small sample inference. Specifically, it will investigate the problems in evaluating saddlepoint approximations for cumulative distribution functions when evaluated near the mean of the distribution approximated, and give alternative formulas for univariate saddlepoint approximation by Lugannani and Rice (1980) and conditional saddlepoint approximation by Skovgaard (1987). It presents a method for constructing confidence regions using the conditional saddlepoint approximations of Skovgaard (1987) for exponential families, and applies saddlepoint methods to calculate the <italic>p</italic>-value using the idea of approximate conditioning proposed by Pierce and Peters (1999). This dissertation also provides a modification to the Skovgaard's conditional saddlepoint approximation for the conditional probabilities in the presence of infinite parameter estimates.
590 $a School code: 0188.
650 4 $a Biology, Biostatistics. $3 1018416
650 4 $a Statistics. $3 517247
690 $a 0308
690 $a 0463
710 2 0 $a The University of Rochester. $3 1249304
773 0 $t Dissertation Abstracts International $g 61-01B.
790 $a 0188
790 1 0 $a Kolassa, John E., $e advisor
791 $a Ph.D.
792 $a 2000
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9959912