Epileptic Seizure Detection based on Hyperparameter Optimization using EEG Data

Epilepsy is a common nerve disorder resulting in seizures. Seizure is best defined as an abnormal electrical activity that can affect brain’s function. EEG Signals are essential for epileptic seizure detection since they show the rise in electrical activity in the brain. Machine learning allows software applications to predict outcomes precisely without being explicitly programmed. Artificial Neural Network (ANN) is the popular classifier in Machine Learning (ML) that categorizes EEG data and detects seizures efficiently without compromising performance. ANN architecture includes several parameters like neurons, activation function, optimizers and so on. For accurate prediction, not all the datasets need the same set of these parameters. So tuning these parameters according to the dataset is important. The proposed work focus on three different methods (Bayes Optimization, Gird Search and Random Search) to search the best values of hyperparameters such as activation function, batch size, epochs, learning rate, number of neurons, and optimizer that support the model to achieve best performance on a given dataset.