Performance Analysis of Different Fuzzy Techniques in Classification of Epilepsy Risk Level for Diabetic Patients using Cerebral Blood Flow,Aggregation Operators and EEG Signals

摘要

This paper compares and analyses the performance of different fuzzy techniques to identify a simple,suitable and reliable fuzzy technique for classification of epilepsy risk level in diabetic patients.The Cerebral Blood Flow(CBF),EEG signal values and aggregation operators are used as parameters.The author's first model is rule based homogeneous fuzzy system with two inputs and twenty-five,rules in rule base reported in Harikumar,et al.The performance and the quality value of this two input homogeneous fuzzy system are low values of 62.56 % and 5.61 respectively.Three different fuzzy techniques with five rules,such as,a two input heterogeneous fuzzy technique,single input rule model(SIRM)and tuned SIRM are analyzed.The performance and quality values obtained for these three fuzzy techniques are 96.94% and 20.62,89.64% and 16.15 and 98.58% and 36.56 respectively.All the above four types of fuzzy techniques are calibrated with data from seventy well-defined diabetic patients and trained with another thirty diabetic patients chosen at random.These systems are tested with another one hundred diabetic patients.The tuned SIRM fuzzy technique is better suited for classification of epilepsy risk level in diabetic patients.