In Epileptic seizure detection, classification of EEG signals are essential measures and allows detecting various causes and symptoms from appropriate EEG signal measures. In this paper, we introduce the design of appropriate Ensemble Empirical Mode Decomposition model with most appropriate parametric measures to accomplish Epileptic event detection in EEG signal with minimized the false rate caused by signal interferences. In addition to this DWT based EEG signal decomposition is also introduced which can categorize the input EEG signal into five different types in accordance with frequency ranges. The proposed model integrates SVM machine learning model for fully automated CAD system with improved classification rate. EEG signal detection and analyzes, on the other hand, uses signal decomposition models for accurate signal detection by smartly rejecting false alarms arise. This system comprised of EMD based IMF band decomposition which requires no human intervention to detect epileptic measures. It allows for prompt accessibility, efficient usage of EEG signal characteristics and provides user convenience. The performance metrics in terms of final classification accuracy and detection rate are experimented with real time data sets extracted from most realistic EEG benchmark epilepsy datasets which considers signal measures from different environmental conditions.