臺閩地區機車交通事故之研究 none

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/43868

Title:	臺閩地區機車交通事故之研究 none
Authors:	郭乃鳳;Nai-feng Kuo
Contributors:	統計研究所
Keywords:	機車肇事;對數線性模型;羅吉斯迴歸;卡方自動互動檢視法;motorcycle traffic accidents;loglinear model;logistic regresstion;CHAID
Date:	2010-07-01
Issue Date:	2010-12-08 14:23:48 (UTC+8)
Publisher:	國立中央大學
Abstract:	臺灣社會因機車的方便性及機動性，已成為一般民眾與中低收入戶的主要代步工具，根據行政院交通部統計，目前機車登記數逐年增加，表示隨著經濟發展及環境開發，民眾對於機車之需求仍持續成長，更因為臺灣在地少人稠，都會區中停車區供不應求情況下，每家每戶平均擁有一部機車，而民眾駕駛機車之違規行為，雖然在法律嚴格執行下，近年來已有所改善，但機車交通事故依然在道路交通事故占有頗高比重，因此國人應該更加重視機車交通事故所帶來的影響。本研究將機車肇事資料依據受傷程分為兩部分，（一）機車駕駛（第一當事人）受傷與否之交通事故方面，利用關聯係數瞭解當事人狀態、事故環境狀況之變數和受傷與否之間相關性及關聯強度，接者透過卡方自動互動檢視挑選重要變數，並且進行二元羅吉斯迴歸模型之分析，（二）機車駕駛於事故發生三十日內死亡之交通事故方面，針對性別、車種、發生時間、保護裝置及飲酒情形等五個變數，以最佳對數線性模型探討死亡事故之特徵，最後針對所討論現象，提出改善及建議。 Because of the convenience and mobility of motorcycle, it has become the popular traffic tool in low and middle-income households in Taiwan. According the report of statistics of Ministry of Transportation, Executive Yuan, the number of registered motorcycles increased annually, with economic and environmental development the masses continue growth in motorcycle demand. Taiwan is a densely populated country, there are shortage of supply of parking spaces in urban area, so that each household in average has at least one motorcycle. Since the motorcycle accidents occupy high proportion of total traffic accidents, we should pay more attention to the impact of motorcycle accidents. For the data set of motorcyclist injured or not in the accident, we calculate the association between response and explanatory variables, and apply CHAID to find the important variables. Further, we establish the logistic regresstion model for those important variables selected in CHAID. For the data set of motorcyclist in accident who die within 30 days, we build up an appropriate loglinear model to find out the characteristic of a fatal accident. Finally, we propose some sugguestions to improve what we observe in this study.
Appears in Collections:	[Graduate Institute of Statistics] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	1460	View/Open

社群 sharing

Loading...