Detection of human physiological state change using Fisher's linear discriminant

This paper describes a new approach to detect physiological state change in human beings through a computer embedded statistical algorithm. Some physiological signals are noisy and therefore changes of such signals are difficult to detect. Also different people generate signals with different characteristics, which makes a simple threshold solution impossible. Fisher's linear discriminant is a widely used statistical algorithm to handle such kind of problems. In this paper we have applied Fisher's discriminant to the detection of physiological state change with some inspiring results. We are designing more structured testing plans to generate data from more subjects and further validate our method(1).