The main aim of this research is to find the feasibility of Linear Discriminant Analysis (LDA) in optimization of fuzzy outputs for epilepsy risk level classification from EEG signals. One of the prominent neurological disorders affecting the nervous system is epilepsy. Due to the hyperactivity of neurons in certain regions of the brain epileptic seizures occur. To classify the epilepsy risk levels based on the parameters extracted like peaks, sharp and spike waves, events, energy, duration, variance and covariance from the EEG signals, the fuzzy pre-classifier is used. The LDA is then applied on the pre-classified data to find exactly the optimized risk levels which clearly describes the epilepsy risk level of the patient. The result analysis show that an average accuracy of 96.152%, an average quality value of 2.098, an average time delay of 2.124 and an average performance index of about 91.9215% is obtained.
Optimization of fuzzy outputs for classification of epilepsy from EEG signals using linear discriminant analysis
01.04.2017
250587 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Nonlinear Discriminant Features Constructed by Using Outputs of Multilayer Perception
British Library Conference Proceedings | 1994
|Probabilistic Linear Discriminant Analysis
British Library Conference Proceedings | 2006
|