Epileptic Seizures: Detection and Classification using Time Series and Deep Learning

Human brain emits signals in the form of Electroencephalogram (EEG) signals. These signals are recorded at regular intervals, therefore, can be represented in the form of Time-Series, with a particular data present at specific time stamp could be very helpful because in case bio signals like that of EEG signals, hitting the nail at its head is very crucial. Feature engineering is one of such techniques that is required to extract the vitals from the EEG data to classify the Epilepsy. Also, it is considered that learning from the past data is of great help, especially to detect the types of epileptic seizures as it provides the insights about the patients’ history and past seizures records. Various libraries of Python like TensorFlow, matplotlib and getML, an Application Programming Interface (API), are used extensively to extract the important features from the EEG data using Time Series Classsification using getML, a Python based API as it uses FastProp, an automatic feature learning algorithm, which makes our work easier and saves our time.