:::

詳目顯示

回上一頁
題名:應用決策樹演算法以探究高科技員工潛在的糖尿病之危險因子
書刊名:健康管理學刊
作者:陳銘樹 引用關係王建智王麗雁
作者(外文):Chen, Ming-shuWang, Chien-chihWang, Li-yan
出版日期:2008
卷期:6:2
頁次:頁135-146
主題關鍵詞:資料探勘決策樹第二型糖尿病危險因子Data miningDecision treeType 2 diabetesRisk factor
原始連結:連回原系統網址new window
相關次數:
  • 被引用次數被引用次數:期刊(1) 博士論文(0) 專書(0) 專書論文(0)
  • 排除自我引用排除自我引用:1
  • 共同引用共同引用:0
  • 點閱點閱:101
高科技產業的產品生命週期短,競爭激烈,雖然員工平均年齡不高,但多數員工超時工作,高度的工作壓力以及缺乏運動或三餐外食的生活習慣促使高科技員工健康逐漸亮起紅燈。在台灣,糖尿病位居十大死因第四位,且有每年上升的趨勢,第二型糖尿病屬於不易自行察覺的疾病,更容易被一般年輕的科技業員工所忽視。本研究以高科技產業員工爲研究對象,根據員工連續兩年的健康檢查報告資料,應用決策樹分類技術分析員工健康狀況的變化,以找出屬於該族群的糖尿病危險因子。過去文獻中的結果,無法讓目前健康指標還在正常範圍內的民眾知道自身健康是否開始產生不好的變化,只能讓健康狀況已經異常的民眾得知自己罹病的機率。因此本研究提出新的分析概念,將「因子變化量」納入考量,如此將可讓民眾知道糖尿病危險因子的變化量是否在安全範圍內。分析結果顯示,考慮因子變化量的決策樹模型預估的準確率最高,總正確率高達82.51%。分析高科技產業員工資料後發現,包含1.空腹血糖;2.血壓;3.三酸甘油脂;4.肝功能(GPT);5.白血球等五項檢驗數據爲第二型糖尿病的統計上的重要危險因子。
The products life cycle of High-Tech industry is short, the competition is intense. Although the employee average age is not high, most employees work overtime, are under high pressure and lack of exercise, employees' lifestyle causes a danger signal to appear. Diabetes has been ranked the fourth position in the top ten causes of death in Taiwan, but has been rising every year. Type 2 Diabetes Mellitus is often difficult to perceive by oneself, and is commonly ignored by young employees in the High-Tech industry. In this study, according to two consecutives years of health examination from employees in the High-Tech industry, applying Decision Tree classification technology analyzes the extent to which the health of staff changes, in order to find out the diabetes risk factors of workers in this industry. The resulting literature is unable to give present employees in the normal range populace an indicator for when their own health starts to take a turn for the worse; rather, to only inform the healthy of levels of the morbidity rate. So this study presentation promotes a new concept that considers the "quantities of factor changes", it will let the populace know if the change quantities of diabetes risk factors are in a safe range or not. Analysis results reveal, assuming the "quantities of factor changes" of the decision tree model have the highest accuracy, the correct rate of 82.51%. After data analyses of the High-Tech industry staff, 5 factors were determined to put one at risk for Type 2 Diabetes Mellitus. These factors are: (1) impaired fasting glucose, (2) blood pressure, (3) triglyceride, (4) Glutamate pyruvate transaminase (GPT), and (5) white blood cell count (WBC).
期刊論文
1.(2000)。科技人健康大調查。康健雜誌。  延伸查詢new window
2.郭育倫、章淑娟(20070900)。100健康亮紅燈 126糖尿病上身--淺談糖尿病前期之健康促進。人醫心傳,45,65-67。  延伸查詢new window
會議論文
1.Kohavi, Ron(1995)。A study of cross-validation and bootstrap for accuracy estimation and model selection。The 14th International Joint Conference on Artificial Intelligence。Morgan Kaufmann。1137-1143。  new window
學位論文
1.廖介銘(2003)。決策樹應用於糖尿病之探勘(碩士論文)。華梵大學。  延伸查詢new window
2.李秀琴(2003)。應用人工智慧技術於人類慢性疾病管理(碩士論文)。彰化師範大學。  延伸查詢new window
3.李文瑞(2004)。運用基因演算法建構疾病早期診斷模型之研究-以糖尿病前期之診斷為例(碩士論文)。輔仁大學。  延伸查詢new window
4.李博智(2002)。資料探勘在慢性病預測模式之建構(碩士論文)。元智大學。  延伸查詢new window
圖書
1.吳宗祐、胡佩怡(2007)。高科技產業從業人員過勞實證研究。臺北:行政院勞工委員會勞工安全衛生研究所。  延伸查詢new window
2.Breiman, L.、Friedman, J. H.、Olshen, R. A.、Stone, C. J.(1984)。Classification and Regression Trees。Chapman & Hall/CRC。  new window
其他
1.1111人力銀行(20080710)。2007科技從業人員心聲大調查,http://www.llll.com.tw/zone/pr/headline.asp?autono=1677。  延伸查詢new window
 
 
 
 
第一頁 上一頁 下一頁 最後一頁 top
:::
無相關書籍
 
無相關著作
 
QR Code
QRCODE