Announcement for Downloading full text filePlease respect the Copyright Act.
All digital full text dissertation and theses from this website are authorized the copyright owners. These copyrighted full-text dissertation and theses can be only used for academic, research and non-commercial purposes. Users of this website can search, read, and print for personal usage. In respect of the Copyright Act of the Republic of China, please do not reproduce, distribute, change, or edit the content of these dissertations and theses without any permission. Please do not create any work based upon a pre-existing work by reproduction, Adaptation, Distribution or other means.
URN etd-0209112-001343 Statistics This thesis had been viewed 3535 times. Download 3973 times. Author Wei-lung Chien Author's Email Address No Public. Department Information Management Year 2011 Semester 1 Degree Master Type of Document Master's Thesis Language zh-TW.Big5 Chinese Page Count 66 Title Prediction of Falling Risk of the Elderly by Data Mining Techniques Keyword Falls in the elderly Classification analysis Feature selection Data mining Risk Prediction Regression Analysis Regression Analysis Risk Prediction Data mining Feature selection Classification analysis Falls in the elderly Abstract A matched case-control (case:control=1:2) study was performed among nursing home residents of 65 years and older during 1-year period. The victims were interviewed with in-charge nurse by questionnaire after the event as soon as possible. Another matched (age±5 year, sex, neighborhood) control group was visited at the same time. All the falls-associated factors such as medication experience, underlying disease, surroundings and the motion accompanied the falls were included as independent variables and the dependent variable was the fall attack.
In this research, the statistical method of Mantel-Haenszel Test is adopted for feature selection and significant variables are preserved. Then 2 data mining techniques, logistic regression and decision tree are applied to construct classification model and the accuracy is compared among intra- variant models.
Instability during the action before the fall is the important factor. The actions including rising from bed or toilet, standing on the chair and taking something from the above-head place, etc. The accuracy of 2 classifiers is all above 80%. The good results can be used to predict falls in the elderly and give reference for the elderly care personnel.
Advisor Committee Yen-ju Yang - advisor
Sung-chien Lin - co-chair
Yu-cheng Kao - co-chair
Files Date of Defense 2012-01-17 Date of Submission 2012-02-09