臺灣博碩士論文加值系統

本研究主要運用人工智慧領域資料探勘（Data Mining），使用醫療健康資料庫資料與空氣污染物監測資料挖掘空氣污染與慢性阻塞性肺疾病（COPD）之關聯性。本研究使用案例交叉（case-crossover）技術設計實驗，並使用條件邏輯斯回歸（Conditional Logistic regression）計算空污對於疾病的勝算比（Odds radios，OR）與95%信賴區間（95% Confidence interval，95% CI）。對2005年至2012年所有因慢性阻塞性肺疾病就診門診資料研究發現，PM2.5、PM10、SO2、NO2、CO污染發生當天及滯後1天慢性阻塞性肺疾病就診門診風險增加，寒冷季節空氣污染對慢性阻塞性肺疾病的影響比溫暖季節更顯著。O3對慢性阻塞性肺疾病的影響存在滯後性，O3污染發生后３至４天內要注意防範。
本研究運用基於集成學習（Ensemble Learning）方法的XGBoost演算法，將當日空污濃度與前一至五天空污濃度作為自變數構建空污與慢性阻塞性肺疾病發病的預測模型。模型預測結果與Random Forest、Neural Network、C5.0、Adaboost及SVM五種算法模型測試結果比較，發現基於集成學習方法XGBoost演算法所構建出的模型在本研究問題上產生出可用性更高的分類結果。

The study aimed to determine whether there is an association between air pollutants levels and outpatient clinic visits with chronic obstructive pulmonary disease (COPD) in Taiwan. Data of air pollutant concentrations (PM2.5、PM10、SO2、NO2、CO、O3) were collected from air monitoring stations. We use a case-crossover study design and conditional logistic regression models with odds ratios (OR) and 95% confidence intervals(CI) for evaluating the associations between the air pollutant factor and COPD-associated OC visits. Analyses show the PM2.5, PM10, CO, NO2, SO2 had significant effects on COPD-associated OC visits. In colder days, a significantly greater effect on COPD-associated OC visits O3 had greater lag effects (the lag was 1， 2，4，5 days) on COPD-associated OC visits. Controlling ambient air pollution would provide benefits to COPD patients.
In this study, We used XGBoost algorithm to build a prediction model of air pollution and hospital readmission for Chrome Obstructive Pulmonary Disease. Compared with Random Forest, Neural Network, C5.0, AdaBoost and SVM, it was found that the model based on the integrated learning method XGBoost algorithm produces a higher classification of this problem result.

摘要 iii
ABSTRACT iv
誌 謝 vi
表目錄 ix
圖目錄 x
第一章、緒論 1
1.1研究背景 1
1.2研究動機 2
1.3研究目的 2
1.4研究流程 3
第二章、文獻探討 5
2.1慢性阻塞性肺疾病 5
2.2慢性阻塞性肺疾病與空污因子 6
2.3資料探勘 7
2.3.1資料探勘的解釋 7
2.3.2資料探勘與醫療相關之應用 8
2.4集成學習 9
2.4.1提升演算法Boosting 9
2.4.2Bootstrap Aggregating（Bagging） 12
2.5全民健保資料庫 13
第三章、研究方法 15
3.1研究資料來源 15
3.2研究對象建置 15
3.3模型測試 17
3.4性能度量 19
3.4邏輯斯回歸 23
3.5梯度提升樹 23
3.6本章小結 27
第四章、研究結果 28
4.1資料描述與清理 28
4.2描述性統計 32
4.2.1門診資料分析 32
4.2.2空污資料分析 33
4.3資料探勘 36
4.4模型建構 35
4.5模型效能評估與比較 37
第五章、結論與建議 37
5.1研究結論 37
5.2研究展望 38
5.3研究限制 38
英文文獻 39
中文文獻 43

英文文獻
Amemiya, Takeshi. Advanced econometrics. Harvard university press, 1985.

Pieter A, Dolf Z., Data Mining. Harlow. Google Scholar, 1996.

Austin, P. C., Tu, J. V., Ho, J. E., Levy, D., & Lee, D. S. (2013). Using methods from the data-mining and machine-learning literature for disease classification and prediction: a case study examining classification of heart failure subtypes. Journal of clinical epidemiology, 66(4): 398-407.

Buda A, Jarynowski A., Life-time Of Correlation And Its Application (volume 1), Wydawnictwo Niezalezne, Wroclaw , 2010

Belleudi V., Faustini A., Stafoggia M., Cattani G., Marconi A., Perucci C. A., & Forastiere F. (2010), Impact of fine and ultrafine particles on emergency hospital admissions for cardiac and respiratory diseases. Epidemiology, 21(3): 414-423.

Berend N. (2016), Contribution of air pollution to COPD and small airway dysfunction. Respirology, 21(2): 237-244.

Berry, Michael J., and Gordon Linoff. Data mining techniques: for marketing, sales, and customer support. John Wiley & Sons, Inc., 1997.

Breiman, L. (1996). Bagging predictors. Machine learning, 24(2): 123-140.

Cai J., Chen R., Wang W., Xu X., Ha S. (2015), & Kan H., Does ambient CO have protective effect for COPD patient?. Environmental research, 136: 21-26.

Cheng M. H., Chiu H. F., & Yang C. Y. (2015), Coarse particulate air pollution associated with increased risk of hospital admissions for respiratory diseases in a tropical city, Kaohsiung, Taiwan. International journal of environmental research and public health, 12(10): 13053-13068.

Ding P. H., Wang G. S., Guo Y. L., Chang S. C., & Wan G. H. (2017), Urban air pollution and meteorological factors affect emergency department visits of elderly patients with chronic obstructive pulmonary disease in Taiwan. Environmental Pollution, 224: 751-758.

Efron, Bradley, and Robert J. Tibshirani. An introduction to the bootstrap. CRC press, 1994.

Fayyad U., Piatetsky-Shapiro G., & Smyth P. (1996), From data mining to knowledge discovery in databases. AI magazine, 17(3): 37.

Friedman, J. H. (2001). Greedy function approximation: a gradient boosting machine. Annals of statistics, 1189-1232.

Gillespie G. (2000), Deploying an IT cure for chronic diseases. Health data management, 8(7): 68-70.

Hwang, S. L., Guo, S. E., Chi, M. C., Chou, C. T., Lin, Y. C., Lin, C. M., & Chou, Y. L. (2016). Association between atmospheric fine particulate matter and hospital admissions for chronic obstructive pulmonary disease in Southwestern Taiwan: a population-based study. International journal of environmental research and public health, 13(4): 366.

Janes H, Sheppard L, Lumley T. (2005), Case–crossover analyses of air pollution exposure data: referent selection strategies and their implications for bias. Epidemiology, 16(6): 717-726.

Kleissner C. Data mining for the enterprise. In: System Sciences, 1998., Proceedings of the Thirty-First Hawaii International Conference on. IEEE, 1998. p. 295-304.

Lee I. M., Tsai S. S., Chang C. C., Ho C. K., & Yang C. Y. (1996), Air pollution and hospital admissions for chronic obstructive pulmonary disease in a tropical city: Kaohsiung, Taiwan. Inhalation toxicology. 19(5): 393-398.

Li X. Y., Gilmour P. S., Donaldson K., & MacNee, W. (1996), Free radical activity and pro-inflammatory effects of particulate air pollution (PM10) in vivo and in vitro. Thorax, 51(12): 1216-1222.

Palaniappan, S., & Awang, R. Intelligent heart disease prediction system using data mining techniques. In: Computer Systems and Applications, 2008. AICCSA 2008. IEEE/ACS International Conference on. IEEE, 2008. p. 108-115.

Safran C., & Chute C. G. (1995), Exploration and exploitation of clinical databases. International journal of bio-medical computing, 39(1): 151-156.

Shaw M. J., Subramaniam C., Tan G. W., & Welge M. E. (2001), Knowledge management and data mining for marketing. Decision support systems, 31(1): 127-137.

Ministry of Health and Welfare, Taiwan Health and Welfare Report 2016, 2016, http://www.mohw.gov.tw/cp-137-521-2.html.

World Health Organization, Tables of health statistics by country , 2016, http://www.who.int/gho/publications/world_health_statistics/2016/en/

World Health Organization, The 10 leading causes of death in the world, 2000 and 2011., 2013.

Tsai S S, Chang C C, Yang C Y. (2013), Fine particulate air pollution and hospital admissions for chronic obstructive pulmonary disease: a case-crossover study in Taipei. International journal of environmental research and public health, 10(11): 6015-6026.

Vestbo J., Hurd S. S., Agustí A. G., Jones P. W., Vogelmeier C., Anzueto A., & Stockley R. A. (2013), Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. American journal of respiratory and critical care medicine, 187(4): 347-365.

Yang C.Y，Chen C.J. (2007), Air Pollution and Hospital Admissions for Chronic Obstructive Pulmonary Disease in a Subtropical City: Taipei, Taiwan. Journal of Toxicology and Environmental Health，Part A: Current Issues,70(14): 1214-1219.

Yorifuji T, Suzuki E, Kashima S.(2014), Hourly differences in air pollution and risk of respiratory disease in the elderly: a time-stratified case-crossover study. Environmental Health, 2014, 13(1): 67.

Yu, W., Liu, T., Valdez, R., Gwinn, M., & Khoury, M. J. (2010). Application of support vector machine modeling for prediction of common diseases: the case of diabetes and pre-diabetes. BMC Medical Informatics and Decision Making, 10(1): 16.

Zou, K.H., O'Malley, A.J., Mauri, L. (2007). Receiver-operating characteristic analysis for evaluating diagnostic tests and predictive models. Circulation, 115(5): 654–7.


中文文獻
黃勝崇，資料探勘應用於醫療院所輔助病患看診指引之研究，南華大學資訊 管理研究所碩士論文，2001.

陳迪祥，以資料探勘技術發掘疾病隱藏關係之研究，國立暨南國際大學資訊管理研究所碩士論文，2003.

李幼平，楊克虎，《循證醫學》，人民衛生出版社，2014.

唐壽生，資料探勘技術應用於肺結核病患完治的預測，國立中正大學資訊管理研究所碩士論文，2004.

朱彩屏，資料探勘在醫療資料庫之研究-以疝氣臨床路徑為例，國立中正大學資訊管理研究所碩士論文，2004.

周志華，《機器學習》，清華大學出版社，2016.

行政院環境保護署，空氣品質指標， http://taqm.epa.gov.tw/taqm/tw/AqiMap.aspx ，2017.

王玉純, 吳俊霖, 莊淳宇, et al. 台灣都會區心血管及呼吸道疾病死亡率與氣象因子之相關分析. 全球變遷通訊雜誌. 2005; 46 14-20.