應用機器學習演算法預測工作者腦心血管疾病風險

Applying Machine-Learning Algorithms to Predict Workers' Risk of Cerebrovascular and Cardiovascular Disease

本研究計畫旨在預測工作者罹患腦心血管疾病的風險，藉由分析工作者的基本資料、體檢結果、個人病史、生活習慣及疲勞量表等，利用機器學習演算法，如×GBoost模型，達成預測。研究結果顯示，該模型對生理指標的預測AUC達99.9%，對心理指標中的睡眠障礙、疲倦及疼痛的預測AUC分別為83%、92%及86%。最終目的是建立健康風險管理機制，提供工作者自我監測與預防指標，達到早期診斷和早期治療，防止過勞引發的腦心血管疾病。

This research project aims to predict the risk of cerebrovascular and cardiovascular diseases in workers by analyzing basic personal information, medical e×amination results, personal medical history, lifestyle habits, and fatigue scales. Machine learning algorithms, such as the×GBoost model, are used for this purpose. The study results show that the model achieves an AUC of 99.9% for predicting physiological indicators and AUCs of 83%, 92%, and 86% for predicting psychological indicators of sleep disorders, fatigue, and pain, respectively. The ultimate goal is to establish a health risk management mechanism that provides workers with self-monitoring and prevention indicators, achieving early diagnosis and treatment to prevent cerebrovascular and cardiovascular diseases caused by overwork.