The Application of Neural Network on the Forecast of Health Tendency of Hospital Staff

• Main Authors: Chung Wen-Feng, 鍾文峰
• Other Authors: Huang Yi-Chao
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/66247676354242033881

碩士 === 國立屏東科技大學 === 工業管理系所 === 97 === With the shift of times and the advance of material life, people's demands of preventive health have been increasing. Hence, regular health examination becomes an important channel for pursuing of health. Furthermore, the influence of habitual behavior and environment makes life style a key to be healthy or not. However, having a good quality of medical care needs to rely on nearly 160,000 medical staff in Taiwan. This study, conducting a questionnaire survey to collect staff's life background in one of the teaching hospital in the south Taiwan and integrating their health examination records. Making the result of questionnaire survey a conditional attribute and health examination a decision attribute one and using Genetic Algorithms to calculate the proportion of them. In addition using Back-propagation Network to predict the correlation of original database insufficiency and completing probability table of disease attributes. The probability value of being ill, which is composed of this correlation; drawing probability graph of correlation between lifestyle and diseases by Bezier surface and using GUI to build forecast system of correlation between personnel's lifestyle and diseases.In conclusion, this system could be used as a reference for hospital staff in daily life and health management.