Composite activity type and stride-specific energy expenditure estimation model for thigh-worn accelerometry

This repository contains code and data for the research project 'Estimation of activity induced energy expenditure using thigh-worn accelerometry and machine learning approaches'.

• The code subfolder contains Jupyter Notebooks and a Python file with helper functions. Further, the models subfolder contains the trained models.
• The data subfolder contains the raw accelerometer files (AX) as well as the raw data from the indirect calorimetry (CPET). Further, different processing files can be found here. The file log_master.csv contains the sociodemographic and timestamp data.
• The figures subfolder contains all relevant figures, which are created as part of running the Jupyter Notebooks. These figures are also part of the research publication.