Spark-Based Deep Learning for Recognition of Epileptic Seizures With Multimodal Brain Waves

Abstract

In this chapter, electrical activity in the brain is measured with the help of brain sensors. Further, electrical activity is stored and analyzed for two different datasets: first, for the epileptic seizure, and second, for brain activity. Next, both epileptic seizure and brain activity datasets are used for deep learning models interfacing with Apache Spark. The deep learning model has a feed-forward neural network, which helps determine features from the epileptic seizure dataset that are important and fit in the hidden patterns among them. Further, FFNN is interfaced with Apache Spark to analyze how it can be beneficial with processing of the deep learning model. The results of the accuracy of deep learning model and processing time caused by Apache Spark help determine how effective the FFNN is for predicting the seizure activity in the brain and how Apache Spark can be applied with a deep learning model to increase its effectiveness for different types of datasets.