Behavior Network-Based Risk Recognition Method

This chapter proposes a two-behavior networks-based method to automatically detect whether a situation is risky or not. Behavior network is used to analyze measured values from the sensors of a smart phone. Bayesian probability is also used for implementing such a behavior network. An experiment was conducted to validate that the speed of recognizing risk situations was improved by learning such situations iteratively through behavior networks with Bayesian probability.