Abstract
we proposed ordinary/unordinary state's label to increase the kind of life log services and to improve the quality of the life log services. We have tried to classify heart rates data to ordinary/unordinary class before. In this paper, we show the results of experiments which we add new characters such as an average of heart rates in a certain window, and time-oriented characters to classify heart rates data with higher accuracy. And we show we classified subjects to two groups. The subjects in one groups, can be esitimated their ordinary/unordinary state from their heart rates. However, we studied the other group's subjects can not be estimated their ordinary/unordinary state from their heart rates. And we observed the case the labeling accuracy getting better with new characters in the case of subjects who can not be estimated their ordinary/unordinary state from their heart rates.
