Sensor-based prediction of physical activity and its impacts using machine learning

This thesis contributed to the development of advanced learning models and multi-sensor decision fusion algorithm to improve the prediction of physical activity and its personal impacts including relative intensity and energy expenditure from wearable sensor data. It identified the optimal sensor positioning and optimal combination of multimodal sensor data for assessing physical activity and predicting its impacts. All methods of this thesis collectively deliver better algorithms and maximise the use of available sensor information to provide accurate measurement of physical activity