What algorithmic approach underlies activity classification in most fitness trackers?. Answer: Convolutional neural networks (CNNs) on accelerometer data.

CNNs are a common approach: they learn patterns in accelerometer signals to classify activities such as walking or running. Trackers may also use other machine-learning methods or combine accelerometer data with signals from other sensors.

Would you also like to know how CNNs classify activities from accelerometer data?AskAnswer complete. One follow-up option available.