A Noise Removal Approach from EEG Recordings Based on Variational Autoencoders
Jamal F. Hwaidi, Tom Chen · 2021
This paper presents a novel approach for reducing noise in electroencephalography (EEG) signals using the Variational AutoEncoders (VAE) algorithm. VAE's are attractive as they are designed on top of standard function approximators neural networks and can be trained with stochastic gradient descent to produce the desired output. Moreover, VAE has been compared with standard fast fixed-point algorithm for independent component analysis (FastICA) algorithm to measure the performance quantitatively by using machine learning algorithms like Support Vector Machines, Naive Bayes, and Decision Tree. The catch of this algorithm is utilised using the concept of misclassification as opposed to the classification accuracy of the above mentioned algorithms.