Art 2-A on T-node machine application to automatic classification of all-night sleep stages
This article presents the parallelisation of the training sequences of the ART 2-A neural network. ART (Adaptive Response Theory) neural models can estimate a decision function due to prototype calculation elements, and it is applied to the automation of the classification of human sleep stages. The study of human sleep consists of monitoring different physiological activities simultaneously and continuously during the night. Each sleep stage can be visually labelled into 30-second periods by an expert. Automation of sleep analysis reduces the computation time by a speed-up factor of 10.
Bibliographic Reference: Paper presented: 2nd International Conference on Research in Computer Science, Ouagadougou (BF), October 12-18, 1994
Availability: Available from (1) as Paper EN 38531 ORA
Record Number: 199411172 / Last updated on: 1994-11-25
Original language: en
Available languages: en