A Study of Practical Causality Acquisition among Vital Signals

According to increase in the number of sensors, the target system could be effectively controlled such as monitor and maintenance. Additionally, transparent causality among sensor signals should be importantly prerequisite for realizing that kind of solution. However, it is very hard to acquire the cause-effect structure among huge number of sensors. In this article, cause-effect structure acquisition is studied and discussed by employing visceral fat estimation of human body as an application. Cause-effect acquisition methods could be mainly classified into two types. One is based on human expertspsila knowledge and the other is using sensory data. They have different effectiveness and ineffectiveness each other. The authors propose the combinational method of them based on the notion of human-machine collaboration.