Abstract:

A daily behavior recognition algorithm based on spectral clustering algorithm and hidden Markov model was proposed by utilizing sensor data of smart phones. After the acceleration, GPS and RSSI data were obtained, our method combined spectral clustering algorithm with hidden Markov model can automatically and effectively identify user behavior. Experimental results show that the proposed method achieved higher recognition accuracy from the reality data.

Key words: smart phone, sensor, activity recognition, hidden Markov model, spectral clustering