政大機構典藏-National Chengchi University Institutional Repository(NCCUR):Item 140.119/68192

政大典藏 > College of Science > Department of Mathematical Sciences > Periodical Articles > Item 140.119/68192

Please use this identifier to cite or link to this item: https://nccur.lib.nccu.edu.tw/handle/140.119/68192

Title:	影響臺灣SARS疫情擴散之因素分析--以類神經網路及迴歸為預測模式
Other Titles:	A Study of Epidemic Outbreak of SARS in Taiwan by Neural Networks and Regression Models
Authors:	陸行;金立人 Luh,Hsing;Chin,Li-Jen
Contributors:	應數系
Keywords:	理論流行病學;倒傳遞網路法;迴歸相關系數;Theoretical Epidemiology;Back-propagation Network;Regression
Date:	2004.12
Issue Date:	2014-08-05 17:04:03 (UTC+8)
Abstract:	The outbreak of the deadly virus of severe acute respiratory syndrome (SARS) grew drastically in multiple ways of spreading from Southern China to many other countries in November 2002. This paper intends to study of the possible factors that relate to the number of SARS infections in Taiwan by Neural Networks and Regression Models. We investigate the number of new SARS cases by considering confirmed cases, suspected cases, probable cases, people with SARS-like symptoms, and number of infections in neighboring countires of Taiwan in that period. By two different study tools with confirmed analysis, we find that there are several important indices to track the number of new case occurrences, including probable and suspected cases, SARS infections and the number of people in quarantine, etc. 2002年十一月中國大陸廣東地區爆發嚴重急性呼吸道症候群(SARS)疫情，疫情迅速漫延世界多國，本文以預防醫學角度提出十種可能影響SARS疫情擴散之因素：可能病例治癒人數、治療中之可能病例數、SARS疑似病例數、死亡病例數、A級居家隔離人數、香港感染人數、中國大陸感染人數、新加坡感染人數、越南感染人數及其他因素，並運用流行病學和醫學統計學等原理與方法，採用類神經網路與迴歸分析兩種研究模式來分析前九種因素對SARS新增感染人數的影響，給予定量分析評價，從兩種方法均可得到相同的結論。由本研究經類神經網路分析和迴歸分析驗證尚在治療中之可能病例數、死亡病例數、居家隔離人數是控制疫情的重要指標。尚在治療中之可能病例數會造成院內感染進而影響感染人數；死亡病例數的公佈能造成民眾更加小心而避免感染到SARS；進行居家隔離能阻絕病源。所以，將來若有類似傳染病時，醫院應該作好防護設施，以及醫護人員應該要有良好的裝備，以避免院內感染的發生。並且應該立即制定良好的隔離政策，對於疫區來的人民、可能或疑似的病人及家屬實施居家隔離。此外公佈病患死亡人數，能使大眾更小心防範，對於控制疫情，也有不錯的效果。
Relation:	智慧科技與應用統計學報,2(2),49-71
Data Type:	article
Appears in Collections:	[Department of Mathematical Sciences] Periodical Articles

Files in This Item:

File	Description	Size	Format
49-71.pdf		948Kb	Adobe PDF2	785	View/Open

社群 sharing

Loading...