© 2020, Springer-Verlag GmbH Germany, part of Springer Nature. Nowadays, Epilepsy is one of the chronic severe neurological diseases; it has been identified with the help of brain signal analysis. The brain signals are recorded with the help of electrocorticography (ECoG), Electroencephalogram (EEG). From the brain signal, the abnormal brain functions are a more challenging task. The traditional systems are consuming more time to predict unusual brain patterns. Therefore, in this paper, effective bio-inspired machine learning techniques are utilized to predict the epilepsy seizure from the EEG signal with maximum recognition accuracy. Initially, patient brain images are collected by placing the electrodes on their scalp. From the brain signal, different features are extracted that are analyzed with the help of the Krill Herd algorithm for selecting the best features. The selected features are processed using an artificial alga optimized general Adversarial Networks. The network recognizes the intricate and abnormal seizure patterns. Then the discussed state-of-art methods are examined simulation results.