Title :
Conditional density estimation for physiological control systems
Author :
Srivastava, Ashok N. ; Weigend, Andreas S.
Author_Institution :
Dept. of Electr. & Comput. Eng., Colorado Univ., Boulder, CO, USA
Abstract :
We introduce a connectionist architecture for predicting the conditional probability density for time series prediction in physiological control systems. The mean of the distribution is the target value, and variance is the confidence in the prediction. The methods are verified on a highly nonlinear system that exhibits deterministic chaos, and a brownian motion process that exhibits multimodal behavior
Keywords :
biocontrol; brownian motion process; cardiovascular system; conditional density estimation; conditional probability density; connectionist architecture; deterministic chaos; highly nonlinear system; multimodal behavior; physiological control systems; target value; time series prediction; Additive noise; Cardiovascular system; Computer architecture; Control systems; Heart rate; Neural networks; Probability density function; Probability distribution; Switches; System identification;
Conference_Titel :
Engineering in Medicine and Biology Society, 1994. Engineering Advances: New Opportunities for Biomedical Engineers. Proceedings of the 16th Annual International Conference of the IEEE
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-2050-6
DOI :
10.1109/IEMBS.1994.415330
