In [8], the authors detect eating activity using a HiddenMarkov Model (HMM) with a wrist-worn sensor. They usebinary classification with eating as one class and all otheractivities as the other class. Similarly, in [9], the authorsrecognize the eating activity using a wrist-worn accelerometerand a gyroscope. A feasibility study on smoking detectionusing a wrist-worn accelerometer is done in [10], where theauthors reported a user-specific accuracy of 70% for thisactivity