Table 4

Examples of items that need to be reported when using machine learning modelling of accelerometer data for PA measurement

Accelerometer placement	Accelerometer initialisation	Machine learning model use
Site	Hardware configuration	Features calculated and extracted
Side of body	Sampling frequency	Outcome variables
Orientation of monitor	Sensor ranges	Access to developed models
Attachment method ▸ Clip ▸ Tape/adhesive ▸ Strap/band	Filtering of raw data ▸ Online or off-line ▸ Design ▸ Order ▸ Corner frequency ▸ Pass-band type	Window length ▸ Overlapping or non-overlapping ▸ Predetermined or based on activity length
		Specific model information and parameter choices