Structured learning for extraction of daily life log measured by smart phone sensors

Abstract

This chapter deals with producing information using structured learning to extract daily life log measured by smart phone sensors. Acceleration, angular velocity, and movement distance are measured by smart phone sensors and stored as the entries of the daily life log together with the activity information and timestamp. First, this chapter introduces the concept of Informationally Structured Space (ISS) and explains how smart phones can be used for collecting data for the daily life log. Then, structured learning is proposed for estimating the human activities based on the time series of the measured data. The method applies growing neural gas for performing clustering on the data, and spiking neural network for estimating the activity. A modified simple spike response model is applied to reduce the computational cost. The external input values of the spiking neurons depend on the growing neurons, and various metrics are investigated in the input calculation. Experiments are performed for verifying the effectiveness of the proposed method. Finally, the future direction on this research is discussed.