基於使用機器學習之患者症狀模型的腹瀉反應和監測系統： 對史瓦帝尼的研究

Response and Surveillance System for Diarrhoea based on a Patient Symptoms Model using Machine Learning: A Study on Eswatini

祖思帛(Zungu, Sibongakonke Kwanele)

指導教授 ： 孫宏民

國立清華大學/電機資訊學院/資訊系統與應用研究所/碩士(2021年)

摘要

基於腹瀉症狀的監測與響應系統，在支持向量機 (SVM) 的幫助下，根據用戶輸入的 6 種症狀預測可能的疾病，輸出是可能發生的疾病。我們也使用了另外兩個模型：隨機森林模型和朴素貝葉斯模型來做比較目的。 此外，根據水資源的地區性缺乏，對可能發生腹瀉爆發的地區進行了預測。 在支持向量機 接收平均值為100%準度，這就是為什麼它主要在系統中使用，其他兩個（隨機森林模型和朴素貝葉斯模型）不同，後者兩個模型在同一數據集上獲得97.62% 的平均準確度。

關鍵字

腹瀉 ； 疾病預測 ； 樸素貝葉斯 ； 隨機森林分類 ； 支持向量機

並列摘要

A surveillance and Response system based on symptoms of diarrhoea, with the help of the Support Vector Machine (SVM) to predict the probable Disease based on 6 symptoms that become an input from the user and the output is the disease which will likely occur. Two other models have been utilized, Random Forest Model and Naïve Bayes Model which are for comparison purposes. Furthermore, a prediction on the area in which a diarrhoea outbreak would likely occur based on the availability or the scarcity of water and the constituency in which the person is giving the symptoms is from. Support vector machine received an average of 100% which is why it will be used in the system unlike the other two (Random Forest Model and Naïve Bayes Model) who received a 97.62% average accuracy on the same dataset.

並列關鍵字

Diarrhoea ； Disease prediction ； Naïve Bayes ； Random Forest Classifier ； Support Vector Machine

參考文獻

Bibliography

[1] Ogbo. A, Agho.K, Ogeleka. P. Woolfenden. S, Page. A, Eastwood. J.(2017). Prevalence and modifying effect of sanitation on infant and young child feeding indicators associated with diarrhoea in sub-Saharan African countries, with high burden of diarrhoea mortality. https://journals.plos.org/plosone/article?id=10.1371/ journal.pone.0171792

[2] Mlilo. P, Mhlanga. B, Senzanje. A. (2020). Emerging Issues in Water Resources Management in Swaziland. http://dlc.dlib.indiana.edu/dlc/bitstream/handle/ 10535/314/Mlilo_139401.pdf?sequence=1

[3] World Health Organization. (2017). Diarrhoeal Disease. http://www.who.int/ news-room/fact-sheets/detail/diarrhoeal-disease.

[4] Maphalala. G, Phungwayo. N, Masona. G, Lukhele. N, Tsegaye. G, Dube. N, Sindisiwe. D, Khumalo. L, Daniel. F, Katsande. R, Tate. E, Mwenda. M, Weldegebriel. G. (2018). Early impact of rotavirus vaccine in under 5year old children hospitalized due to diarrhea, Swaziland. Vaccine https://pubmed.ncbi.nlm.nih.gov/28778615/