An appropriate number of neurons in a hidden layer for personal authentication using delta brainwave signals

This study takes a further look into the appropriate number of neurons in a hidden layer for personal authentication which uses delta brainwave signals. In line with the principle of supervised neural network, the number of neurons in the hidden layer is an important factor to make learning more effective. The objective of this research paper is to study the number of neurons in the hidden layer. Accordingly, 1000 data points of electroencephalogram (EEG) signals in a group of four channels, F4, P4, C4, and O2, are explored. The practical technique, Independent Component Analysis (ICA) by SOBIRO algorithm, which is considered as a clean algorithm, separates the individual signals from background noise. Delta brainwaves are extracted from brain signal for identifying 30 subjects, by supervised neural network. The number of neurons in the hidden layer (1-30 neurons) is used to test the accuracy of the identifying process. The information will then be classified into 20 to 30 subjects to find the appropriate number of neurons in the hidden layer in each group.