OGCP: Experimental Evaluation of Epilepsy Disease Prediction Using Optimized Gradient Classification Principle
H. Prasad, G. Ramkumar
Abstract
Epilepsy is a neurological disorder characterized by recurrent seizures, and accurate, real-time prediction of seizures is critical for timely interventions. This study presents an Optimized Gradient Classification Principle (OGCP) with a Hybrid Deep Learning (DL) Model for predicting epilepsy from EEG signals. The proposed model was evaluated using the Epilepsy Seizure dataset from Kaggle and compared against nine well-established models, including CNN, LSTM, GRU, ResNet50, XGBoost, and SVM. The performance of the models was evaluated based on key metrics such as accuracy, precision, F1-score, recall, inference time, and model size. The proposed model outperformed all other models, achieving an impressive accuracy of 96.89%, precision of 94.56%, and F1-score of 95.78<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>. The model's inference time was 43.4 ms, making it suitable for real-time applications. Additionally, the model's relatively moderate size of 57.9 MB ensures efficient deployment in memory-constrained devices. These results demonstrate the superiority of the OGCP-based Hybrid DL model for epilepsy prediction, making it a promising approach for medical applications where accuracy and real-time performance are crucial. Future work will focus on improving inference time and incorporating additional physiological signals to enhance predictive performance.