Residual Neural Networks to Distinguish Craving Smokers, Non-craving Smokers and Non-smokers by their EEG signals

Abstract

We investigate the differences in brain signals of craving smokers, non-craving smokers, and non-smokers. To this end, we use data from resting-state EEG measurements to train predictive models to distinguish these three groups. We improve the neural network models applied earlier in two ways: firstly by adding channel-wise convolutional layers, secondly by adding residual connections to the network. We further extend the validation to make it similar to a real world scenario, in which a prediction is based on all data available for this measurement. Finally, we analyze the prediction quality for each measurement individually. Our results demonstrate significant improvements.