Machine learning for the analysis of healthy lifestyle data : a scoping review protocol

Abstract

The objective of this scoping review is to identify and characterize machine learning algorithms used in data analysis of healthy lifestyle. The specific objectives are the study of a) terminology, b) healthy lifestyle variables analysed either input or output, c) programs and libraries used to analyse data, and d) sources, types, and quality of data analysed. Eligibility criteria: In this scoping review the inclusion criteria from studies that provide empirical information are as follows: a) studies must use machine learning models either supervised or unsupervised learning to analyses lifestyle data (input or output), b) studies must use real data from individuals for analysis, and c) the language of the studies must be English or Spanish. Furthermore, theorical studies focusing on a) the mathematical approach to explain algorithm construction, and b) guidelines for implementing the use of machine learning in the field of health will be excluded. Additionally, studies with simulated data and those aiming to develop a robot or app based on machine learning will be excluded. Regarding lifestyle, studies whose main topic is substance use, such as alcohol intake, or those related to smoking cessation will be manually excluded.