Deep Learning for Neuromarketing; Classification of User Preference using EEG Signals

The present study investigates the applicability of deep learning methods in EEG neuromarketing prediction tasks, compared to traditional machine learning approaches. Neuroscientific methods have expanded research capabilities in marketing and created new insights into consumer behavior and decision making processes. Both machine learning and deep learning approaches can be employed to predict relevant consumer preference from brain activity. The former requires extensive signal processing and feature engineering for classification whereas the later relies on raw brain signals and thus avoids time-consuming preprocessing. In this paper, the performance of a machine learning model comprising an ensemble of algorithms was compared to the performance of a convolutional neural network (CNN) on two independently collected EEG datasets, one concerning product choices and the other movie ratings. While both models showed poor performance for prediction of product choices, the convolutional neural network proved more accurate in the prediction of movie ratings. This provides evidence for the superiority of deep learning algorithms in certain neuromarketing prediction tasks. We discuss the limitations and future application opportunities.