A Comprehensive Analysis on Post-processing Mathematical Models (MRE, Aggregation Operators and Soft Decision Trees) for Patient Specific Fuzzy based Epilepsy Risk Level Classifier from EEG Signals

摘要

The purpose of this paper is to identify a comprehensive post processing mathematical models such as Minimum Relative Entropy (MRE), Aggregation operators, and Soft Decision Tree (SDT) in optimization of epilepsy risk levels obtained from fuzzy classifier. The fuzzy pre-classifier is used to classify the risk levels of epilepsy based on extracted parameters like energy, variance, peaks, sharp and spike waves, duration, events and covariance from the EEG signals of the patient. MRE, aggregation operator, and SDT (post-classifier with max-min criteria -six types) are applied on the classified data to identify the optimized risk level (singleton) which characterizes the patient's epilepsy risk level. The efficacy of the above methods is compared based on the bench mark parameters, such as, performance index (PI), New Measures and quality value (Q V). A group of ten patients with known epilepsy findings are used for this study High PI such as 95.88 % was obtained at QV's of 22.43 in the all three groups of post processing methods, when compared to the value of 40% and 6.25 through fuzzy classifier respectively. With the in sight of new measures and minimum overheads, it is identified that the Soft Decision Trees (agg and max-min) method is accounted as a compromise post-classifier.