Approximate entropy and support vector machines for electroencephalogram signal classification.

The article presents a study which confirms whether approximate entropy waves can be effectively used to the automatic real-time epilepsy detection in the electroencephalogram, and explores its generalization ability as a classifier trained utilizing a nonlinear dynamics index. The study used a support vector machine and a nonlinear dynamics index-approximate entropy. Results showed that the combination of machines and entropy reveals good generalization ability.