According to the World Health Organization (WHO), “Epilepsy” is one of the most widespread and common neurological disorders worldwide. Prediction of this condition through early seizure technique plays a crucial role in improving the health condition of the individuals and patients suffering from epileptic criteria in their day-to-day lives. Automated neurological systems (which classify and detect these criteria automatically through machine learning) typically rely on raw forms of EEG data signals as inputs, this approach helps in minimizing and reducing of the processing power and computational load needed for categorizing data. Machine learning models have proven effective, hence are recommended for detecting epileptic events, as they show a significant contribution to patient diagnosis while it also enhances prediction speed and classification accuracy of the given data. This model uses algorithms such as: Random Forest (RF), Fuzzy Rough Nearest Neighbor (FRNN), k-Nearest Neighbor (KNN), and Support Vector Machine (SVM) algorithms, which have advantages in prediction of epileptic seizure development earlier compared to the reference model. This proposed framework classifies an unidentified EEG signal by segmenting it into ictal and interictal phases. This model was evaluated using experimental testing on two well-known standard datasets, naming: The Bonn dataset and The Children’s Hospital of Boston-Massachusetts Institute of Technology (CHB-MIT) datasets. The findings depicts that, across both datasets, Fuzzy Rough Nearest Neighbor (FRNN) and K-Nearest Neighbor (KNN) achieve the classification accuracy in the highest form, as well as enhanced specificity and sensitivity rates.