Using Mega-Trend-Diffusion to Verify the Effectiveness of Psychiatric Disorders by Integrating Western and Traditional Medicine Therapy

• Main Authors: Yuk-HingNg, 吳郁馨
• Other Authors: Der-Chiang Li
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/67361278127905748380

碩士 === 國立成功大學 === 工業與資訊管理學系碩士在職專班 === 102 === The research is accomplished by using therapy to mental patients (Neuro-Psychiatric patients) integrated traditional Chinese Herb Science and modern Western medical science to decrease the possibility of potential reaction or side effect from medicine. The experimental data collected from patients under the integrated therapy treatment in a mental health care institution in Southern Taiwan in the interval of three years from 2010 to 2012, the significant samples collected were only 29 which is not sufficiently to achieve stable results with Back-Propagation Neural Network modeling tools in learning. So, this research intends to generate virtual samples by Mega-Trend-Diffusion. This may help to stabilize the Back-Propagation Neural Network learning process. Experimental results show the stability and classification correction rate are higher than the original samples used in this research. In the follow-up research adding more variables and the number of samples, developing better way to generate virtual samples. We hope this model may provide direction to relate research of the integration of both traditional Chinese Herb Science and the modern Western medical science in treatment of mental illness having alternative therapy information.