A Survey of Machine Learning algorithms in EEG

This research focuses on surveying machine learning algorithms and providing a suitable model to match electroencephalogram (EEG) signal classification needs. Data is recorded by self-written an application, including 11 states: the states related to eye behavior, facial expression, and thinking signals. Feature extraction is an essential step in the process of signal, and the Wavelet transform is used in this phase. The feature matrix is built with components calculated by using the Wavelet method. This paper also gives a piece of information about classical machine learning such as Support Vector Machine (SVM) and Bagged Tree (BT) and modern algorithms as Artificial Neural Network (ANN) and Deep Learning (DL). There are four main algorithms used to classify signals. The proposed scheme yielded significantly higher classification performances using machine learning classifiers compared to extant quantitative feature extraction. These results suggest the proposed feature extraction method reliably classifies EEG signals recorded during thinking with a higher degree of accuracy.