Identification of Smartwatch-Collected Lifelog Variables Affecting Body Mass Index in Middle-Aged People Using Regression Machine Learning Algorithms and SHapley Additive Explanations

Abstract

Body mass index (BMI) plays a vital role in determining the health of middle-aged people, and a high BMI is associated with various chronic diseases. This study aims to identify important lifelog factors related to BMI. The sleep, gait, and body data of 47 middle-aged women and 71 middle-aged men were collected using smartwatches. Variables were derived to examine the relationships between these factors and BMI. The data were divided into groups according to height based on the definition of BMI as the most influential variable. The data were analyzed using regression and tree-based models: Ridge Regression, eXtreme Gradient Boosting (XGBoost), and Category Boosting (CatBoost). Moreover, the importance of the BMI variables was visualized and examined using the SHAPley Additive Explanations Technique (SHAP). The results showed that total sleep time, average morning gait speed, and sleep efficiency significantly affected BMI. However, the variables with the most substantial effects differed among the height groups. This indicates that the factors most profoundly affecting BMI differ according to body characteristics, suggesting the possibility of developing efficient methods for personalized healthcare.

Machine learning analysis and risk prediction of weather-sensitive mortality related to cardiovascular disease during summer in Tokyo, Japan

“…6 . The Boruta feature selection is a wrapper method for detecting important features in ML 44 , 45 : Shuffled duplicates (shadow features as noise) of all features are added as unpredictability to the original feature dataset (e.g., T max , T max Pre 2, …, Pres , PresPre 2, …); next, feature importance based on Z-scores in the enlarged dataset (i.e., original features + shadow features) is used to train a decision tree-based algorithm (Gradient Boosting Decision Trees in this study). Each training cycle is analysed for a higher priority feature than the most important shadow feature, and elements considered highly irrelevant are deleted.…”

Detecting Need-Attention Patients using Machine Learning

“…The researchers recommended Boruta to select relevant variables in high-dimensional datasets [37], [38]. However, [31] stated that it is difficult to identify the variables that are close to their best shadow features in Boruta algorithm.…”

Machine learning analysis and future risk prediction of weather-sensitive cardiovascular disease mortality during summer in Tokyo, Japan

“…6. The Boruta feature selection is known as a wrapper method for detecting important features in the ML eld 43,44 : Shu ed duplicates (shadow features as "noise") of all features are added as unpredictability to the original feature dataset (T max , T max Pre2, …, Pres, PresPre2, …, etc. ); then, feature importance based on Z-score on the enlarged dataset (i.e., original features + shadow features) is used to train a decision tree-based algorithm (Gradient Boosting Decision Trees adopted in this study).…”