User-Independent Activity Recognition Method Using Genetic Algorithm based Feature Selection

摘要

A method for recognizing a user-independent behavior according to the present invention comprises the steps of: (a) carrying a mobile terminal equipped with an accelerometer for a plurality of target users and performing a predetermined number of actions, Collecting data; (b) extracting, from the collected data, a plurality of characteristics for each behavior for each of a plurality of target users; (c) using the genetic algorithm to select some of the plurality of target users of the plurality of features and some features important to the plurality of behaviors; And (d) classifying the user's behavior using the selected features.;