Health status prediction for the elderly based on machine learning. (September 2020)

• Record Type: Journal Article
• Authors: Qin, Fang-Yu; Lv, Zhe-Qi; Wang, Dan-Ni; Hu, Bo; Wu, Chao
• Is Part Of: Archives of gerontology and geriatrics. Volume 90(2020)
• Publication Date: 2020-09
• Subjects: Machine learning -- Elderly -- Health prediction -- Social service -- Data-driven; Aging -- Periodicals; Geriatrics -- Periodicals; Gerontology -- Periodicals; Electronic journals; 305.26
• DOI: 10.1016/j.archger.2020.104121
• Languages: English
• ISSNs: 0167-4943

Abstract

Highlights: The machine learning methods help researchers select the predictors of health status in the older population efficiently. The machine learning methods automatically capture the complicated relationships between the non-linear predictors and the health outcomes. The artificial neural networks have the best prediction accuracy in relation to older people's self-reported health. Abstract: Health and social care services are crucial to old people. The provision of services to the elderly with care needs requires more accurate predictions of the health status of the elderly to rationalize the allocation of the limited social care resources. The traditional analytical methods have proved incapable of predicting the demands of today's society, compared to which machine learning methods can more accurately capture the nonlinear relationships between the variables. To ascertain visually the performance of these machine learning methods regarding the prediction of the elderly's care needs, we designed and verified the experiment.