@Matthew_Ahmadi_ , @NKoemel latest work demonstrates the capacity of machine learning to harmoniously classify activities between wrist- and thigh-worn devices. They achieved >90% classification accuracy; a key step toward combining studies! See more:⌚️link.springer.com/article/10.118…
link.springer.com Development and evaluation of wrist- and thigh-worn accelerometer algorithms using self-training... International Journal of Behavioral Nutrition and Physical Activity - Wearable accelerometers are widely used in health research, but differing placements (e.g. wrist vs. thigh) hinder harmonising... 