Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients
碩士 === 國立雲林科技大學 === 工業工程與管理系 === 106 === Burns are one of the major injuries in Taiwan. When burns occur unexpectedly, they often accompany longer hospital stays. According to the 2016 statistics of the Ministry of Health and Welfare, the average length of stay in special wards was highest at 11.49...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2018
|
Online Access: | http://ndltd.ncl.edu.tw/handle/q7953e |
id |
ndltd-TW-106YUNT0031062 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-106YUNT00310622019-05-16T00:44:54Z http://ndltd.ncl.edu.tw/handle/q7953e Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients 應用人工免疫演算法於燒燙傷病人住院時間預測 Yu-Sheng Chen 陳昱昇 碩士 國立雲林科技大學 工業工程與管理系 106 Burns are one of the major injuries in Taiwan. When burns occur unexpectedly, they often accompany longer hospital stays. According to the 2016 statistics of the Ministry of Health and Welfare, the average length of stay in special wards was highest at 11.49 days except for chronic and sub-acute respiratory care beds. Burn patients had repeat reconstruction operations and rehabilitation programs, and they will affect the patient's mental health in hospital. Therefore, effective management of bed resources and planning of better inpatient medical services are necessary. The study was used the Data from the Taiwanese National Health Insurance Research Databases. In our study, we used the methods of Artificial Immune System (AIS) to combine the Back Propagation Neural Network (BPN) and k-NN and Support Vector Machine (SVM) to construct the three forecast models. At the last, we use the Friedman's test and T-test to evaluate the performance of the forecast models. The accuracy of AIS combine BPN forecast model and the AUC individual is 92.66% and 0.88. The accuracy of AIS combine SVM forecast model and the AUC individual is 92.95% and 0.888. Both models are better than the AIS combine k-NN. Therefore, the forecast models that AIS combine BPN and SVM are suitable for the length of hospital stay of the burn patients. And help the doctors and hospitals take the different ways to treat the different burn patients, and improve medical resources management and service planning to reduce the medical costs. Bor-Wen Cheng 鄭博文 2018 學位論文 ; thesis 54 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立雲林科技大學 === 工業工程與管理系 === 106 === Burns are one of the major injuries in Taiwan. When burns occur unexpectedly, they often accompany longer hospital stays. According to the 2016 statistics of the Ministry of Health and Welfare, the average length of stay in special wards was highest at 11.49 days except for chronic and sub-acute respiratory care beds. Burn patients had repeat reconstruction operations and rehabilitation programs, and they will affect the patient's mental health in hospital. Therefore, effective management of bed resources and planning of better inpatient medical services are necessary.
The study was used the Data from the Taiwanese National Health Insurance Research Databases. In our study, we used the methods of Artificial Immune System (AIS) to combine the Back Propagation Neural Network (BPN) and k-NN and Support Vector Machine (SVM) to construct the three forecast models.
At the last, we use the Friedman's test and T-test to evaluate the performance of the forecast models. The accuracy of AIS combine BPN forecast model and the AUC individual is 92.66% and 0.88. The accuracy of AIS combine SVM forecast model and the AUC individual is 92.95% and 0.888. Both models are better than the AIS combine k-NN.
Therefore, the forecast models that AIS combine BPN and SVM are suitable for the length of hospital stay of the burn patients. And help the doctors and hospitals take the different ways to treat the different burn patients, and improve medical resources management and service planning to reduce the medical costs.
|
author2 |
Bor-Wen Cheng |
author_facet |
Bor-Wen Cheng Yu-Sheng Chen 陳昱昇 |
author |
Yu-Sheng Chen 陳昱昇 |
spellingShingle |
Yu-Sheng Chen 陳昱昇 Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients |
author_sort |
Yu-Sheng Chen |
title |
Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients |
title_short |
Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients |
title_full |
Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients |
title_fullStr |
Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients |
title_full_unstemmed |
Application of Artificial Immune System Algorithm to Predict the Length of Hospital Stay of Burn Patients |
title_sort |
application of artificial immune system algorithm to predict the length of hospital stay of burn patients |
publishDate |
2018 |
url |
http://ndltd.ncl.edu.tw/handle/q7953e |
work_keys_str_mv |
AT yushengchen applicationofartificialimmunesystemalgorithmtopredictthelengthofhospitalstayofburnpatients AT chényùshēng applicationofartificialimmunesystemalgorithmtopredictthelengthofhospitalstayofburnpatients AT yushengchen yīngyòngréngōngmiǎnyìyǎnsuànfǎyúshāotàngshāngbìngrénzhùyuànshíjiānyùcè AT chényùshēng yīngyòngréngōngmiǎnyìyǎnsuànfǎyúshāotàngshāngbìngrénzhùyuànshíjiānyùcè |
_version_ |
1719170523291713536 |