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

Classification with high-dimensional...

Huang, Jing.

FindBook

Google Book

Amazon

博客來

Classification with high-dimensional predictors and qualitative response. Applications to genetics.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Classification with high-dimensional predictors and qualitative response. Applications to genetics./
作者:	Huang, Jing.
面頁冊數:	91 p.
附註:	Adviser: Richard A. Olshen.
Contained By:	Dissertation Abstracts International63-04B.
標題:	Biology, Biostatistics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3048543
ISBN:	0493628746

Classification with high-dimensional predictors and qualitative response. Applications to genetics.
Huang, Jing.

Classification with high-dimensional predictors and qualitative response. Applications to genetics. - 91 p.

Adviser: Richard A. Olshen.

Thesis (Ph.D.)--Stanford University, 2002.

In many real-world situations, multiple factors combine to determine outcome; those factors can have complicated and influential interactions, though perhaps small and insignificant individual contributions. One good example of such a situation is the association between polygenic disease and many genetic and environmental risk factors. Traditional approaches focusing on studying effects of individual factors have proven to be inadequate for such situations. My approach, on the contrary, does treat all risk factors together. It considers both interactions and complex combined effects simultaneously. The technique developed in this thesis stems from the traditional rooted binary classification tree method. It uses the binary tree as framework and employs penalized linear regression to define the specific rule for partitioning. Join classification tree and regression together not only allows us to consider combined effects and interactions simultaneously, but also provides us a simple, easy-to-interpret model. In addition, its predictive power and robustness are improved by the variable selection technique used in the binary split. This technique comprises several known basic components: cross-validation, bootstrapping, and permutation testing. The combination is novel. This algorithm has been applied to simulated data and also to data from a large scale genetic study. It demonstrates substantial improvement in performance over prior competitive binary tree-structured methodologies. Theoretical consistency of this method is also discussed.

ISBN: 0493628746Subjects--Topical Terms:

1018416
Biology, Biostatistics.

Classification with high-dimensional predictors and qualitative response. Applications to genetics.
LDR:02486nam 2200289 a 45 001 936753
005 20110510
008 110510s2002 eng d
020 $a 0493628746
035 $a (UnM)AAI3048543
035 $a AAI3048543
040 $a UnM $c UnM
100 1 $a Huang, Jing. $3 1036865
245 1 0 $a Classification with high-dimensional predictors and qualitative response. Applications to genetics.
300 $a 91 p.
500 $a Adviser: Richard A. Olshen.
500 $a Source: Dissertation Abstracts International, Volume: 63-04, Section: B, page: 1912.
502 $a Thesis (Ph.D.)--Stanford University, 2002.
520 $a In many real-world situations, multiple factors combine to determine outcome; those factors can have complicated and influential interactions, though perhaps small and insignificant individual contributions. One good example of such a situation is the association between polygenic disease and many genetic and environmental risk factors. Traditional approaches focusing on studying effects of individual factors have proven to be inadequate for such situations. My approach, on the contrary, does treat all risk factors together. It considers both interactions and complex combined effects simultaneously. The technique developed in this thesis stems from the traditional rooted binary classification tree method. It uses the binary tree as framework and employs penalized linear regression to define the specific rule for partitioning. Join classification tree and regression together not only allows us to consider combined effects and interactions simultaneously, but also provides us a simple, easy-to-interpret model. In addition, its predictive power and robustness are improved by the variable selection technique used in the binary split. This technique comprises several known basic components: cross-validation, bootstrapping, and permutation testing. The combination is novel. This algorithm has been applied to simulated data and also to data from a large scale genetic study. It demonstrates substantial improvement in performance over prior competitive binary tree-structured methodologies. Theoretical consistency of this method is also discussed.
590 $a School code: 0212.
650 4 $a Biology, Biostatistics. $3 1018416
650 4 $a Biology, Genetics. $3 1017730
650 4 $a Statistics. $3 517247
690 $a 0308
690 $a 0369
690 $a 0463
710 2 0 $a Stanford University. $3 754827
773 0 $t Dissertation Abstracts International $g 63-04B.
790 $a 0212
790 1 0 $a Olshen, Richard A., $e advisor
791 $a Ph.D.
792 $a 2002
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3048543