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 到查詢結果 [ null ]

切換: 標籤 | MARC模式 | ISBD

Exploring the computational and impl...

Dai, Dajun.

FindBook

Google Book

Amazon

博客來

Exploring the computational and implementation characteristics of an improved genetic algorithm for cluster detection.

紀錄類型:	書目-語言資料,印刷品 : Monograph/item
正題名/作者:	Exploring the computational and implementation characteristics of an improved genetic algorithm for cluster detection./
作者:	Dai, Dajun.
面頁冊數:	150 p.
附註:	Adviser: Tonny J. Oyana.
Contained By:	Dissertation Abstracts International68-05A.
標題:	Artificial Intelligence. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3264806
ISBN:	9780549034704

Exploring the computational and implementation characteristics of an improved genetic algorithm for cluster detection.
Dai, Dajun.

Exploring the computational and implementation characteristics of an improved genetic algorithm for cluster detection. - 150 p.

Adviser: Tonny J. Oyana.

Thesis (Ph.D.)--Southern Illinois University at Carbondale, 2007.

Cluster detection is an important exploratory data-mining tool in Geographic Information Science (GISci). Traditional cluster detection methods in GISci assume clusters to be in either circular or in horizontal and vertical elliptic shapes. However, these methods are limited and may not capture diagonally elongated clusters. Consequently, these methods often delineate clusters with low accuracy, take lengthy runtime, or at times fail to report clusters.

ISBN: 9780549034704Subjects--Topical Terms:

769149
Artificial Intelligence.

Exploring the computational and implementation characteristics of an improved genetic algorithm for cluster detection.
LDR:03109nam 2200325 a 45 001 958911
005 20110704
008 110704s2007 ||||||||||||||||| ||eng d
020 $a 9780549034704
035 $a (UMI)AAI3264806
035 $a AAI3264806
040 $a UMI $c UMI
100 1 $a Dai, Dajun. $3 1282376
245 1 0 $a Exploring the computational and implementation characteristics of an improved genetic algorithm for cluster detection.
300 $a 150 p.
500 $a Adviser: Tonny J. Oyana.
500 $a Source: Dissertation Abstracts International, Volume: 68-05, Section: A, page: 2110.
502 $a Thesis (Ph.D.)--Southern Illinois University at Carbondale, 2007.
520 $a Cluster detection is an important exploratory data-mining tool in Geographic Information Science (GISci). Traditional cluster detection methods in GISci assume clusters to be in either circular or in horizontal and vertical elliptic shapes. However, these methods are limited and may not capture diagonally elongated clusters. Consequently, these methods often delineate clusters with low accuracy, take lengthy runtime, or at times fail to report clusters.
520 $a The goal of this dissertation is to develop an improved genetic algorithm (GA) with a new flexible gene structure for detecting clusters in both circular and elongated clusters in any direction. By introducing flexibly oriented ellipses as the gene structure to better represent diagonally elongated clusters, a candidate solution list to record all possible solutions, and a combination algorithm to merge overlapping solutions, the improved genetic algorithm in this research delineates cluster boundaries close to the real shapes of clusters. It also has a competent runtime and a high percentage rate of cluster detection. The performance test on both synthetic and real-world data sets with the comparison to PROgram for CLUster DEtection and Kulldorff's suggests that the improved genetic algorithm in this research can be a viable accurate, efficient, and reliable approach to cluster detection.
520 $a The combination of the improved genetic algorithm with an artificial neural network method, self-organizing maps (SOMs) shows that improved genetic algorithm can be effective and reliable in high-dimensional environmental data mining. Finally, using the improved genetic algorithm, combined with SOMs, Geographic Information Systems (GIS), spatial techniques and statistical methods, this research evaluates dioxin contamination in the city of Midland and the Tittabawassee River floodplain in Michigan and its impact on the prevalence rates of adult female breast cancer. The improved genetic algorithm delineates the clusters of breast cancer accurately and at the same time minimizes the inclusion of background population at risk.
590 $a School code: 0209.
650 4 $a Artificial Intelligence. $3 769149
650 4 $a Environmental Sciences. $3 676987
650 4 $a Geography. $3 524010
650 4 $a Health Sciences, Epidemiology. $3 1019544
690 $a 0366
690 $a 0766
690 $a 0768
690 $a 0800
710 2 $a Southern Illinois University at Carbondale. $b Environmental Resources & Policy. $3 1282377
773 0 $t Dissertation Abstracts International $g 68-05A.
790 $a 0209
790 1 0 $a Oyana, Tonny J., $e advisor
791 $a Ph.D.
792 $a 2007
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3264806