Genetic fuzzy classifier for sleep stage identification
Computers in Biology and Medicine
Expert Systems with Applications: An International Journal
LIBSVM: A library for support vector machines
ACM Transactions on Intelligent Systems and Technology (TIST)
Multi-scale stacked sequential learning
Pattern Recognition
A comparison of methods for multiclass support vector machines
IEEE Transactions on Neural Networks
| Hi-index | 0.00 |
Manual sleep stage classification is a tedious process that takes a lot of time to sleep experts performing data analysis or studies on this field. Moreover errors and inconsistencies between classifications of the same data are frequent. Due to this, there is a great need of automatic classification systems to support reliable classification. This work extends the work by Herrera et al. (International Journal of Neural Systems 10.1142/S0129065713500123), inspecting the use of two techniques to improve the accuracy of sleep stage classifiers based on support vector machines from electroencephalogram, electrooculogram and electromyogram signals. Moreover, three different support vector machine multi-classifiers have been tested to evaluate and compare their performance. To accomplish these tasks, three different feature extraction techniques are applied to the electroencephalogram signals. First, the joint use of these feature sets, together with the electrooculogram and electromyogram information, is inspected (and compared with the use of each feature extraction method separately). Second the possibility of using nearby stages information to predict the current stage is inspected. Results obtained show significant improvements in the classification rates achieved using the two proposed techniques.