Abstract
With the aging of the global population, the number of people with chronic diseases is also increasing, and cardiac disease has become the main cause of human deaths worldwide. In this study, we propose an integrated detection system for measuring the blood pressure (BP), blood glucose (BG), blood lipids (BL), and heart rate (HR). Next, we employ five commonly used machine-learning-based (ML-based) data classification methods, namely, support vector machine (SVM), random forests (RF), k-nearest neighbors (KNN), XGBoost, and LightGBM, for predicting chronic cardiac disease. These five classification methods use the data of BP, BG, BL, and HR, to predict the chronic cardiac disease, whose result shows that RF and KNN have the highest prediction accuracy (88.52%) as compared to the new ML methods, such as XGBoost and LightGBM. In addition, the proposed system should serve as a platform for the long-term detection and tracking of users’ physical health.
Keywords
Chronic disease; Cardiac disease; Machine learning; Data analysis
DOI
https://doi.org/10.21606/drs.2020.351
Citation
Chang, C., Wu, Y., Yang, C., Wu, M., Wu, T., Liu, Y., Chen, C., and Lin, Y. (2020) A Health Care Platform Design: Applying Novel Machine Learning Methods to Predict Chronic Cardiac Disease, in Boess, S., Cheung, M. and Cain, R. (eds.), Synergy - DRS International Conference 2020, 11-14 August, Held online. https://doi.org/10.21606/drs.2020.351
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
A Health Care Platform Design: Applying Novel Machine Learning Methods to Predict Chronic Cardiac Disease
With the aging of the global population, the number of people with chronic diseases is also increasing, and cardiac disease has become the main cause of human deaths worldwide. In this study, we propose an integrated detection system for measuring the blood pressure (BP), blood glucose (BG), blood lipids (BL), and heart rate (HR). Next, we employ five commonly used machine-learning-based (ML-based) data classification methods, namely, support vector machine (SVM), random forests (RF), k-nearest neighbors (KNN), XGBoost, and LightGBM, for predicting chronic cardiac disease. These five classification methods use the data of BP, BG, BL, and HR, to predict the chronic cardiac disease, whose result shows that RF and KNN have the highest prediction accuracy (88.52%) as compared to the new ML methods, such as XGBoost and LightGBM. In addition, the proposed system should serve as a platform for the long-term detection and tracking of users’ physical health.