Unveiling the secrets of online consumer choice: a deep learning algorithmic approach to evaluate and predict purchase decisions through EEG responses

This study utilized cognitive neuroscience experiments to assess and predict online individual behavior by evaluating brain activity signals. We conducted an event-related potential (ERP) experiment and analyzed the data obtained from 85 participants. Moreover, we employed a deep learning algorithm to predict purchase decision-making behavior by examining four ERP components as predictive indicators. Empirical results indicated that presentation order effects were induced when participants perceived different presentation orders of three decision support tools. Importantly, the experimental results revealed an accuracy and F1-score of 98% and 0.98, respectively, for consumers’ choice prediction using a convolutional neural network (CNN). Our study not only ushered in a new data collection scheme for information system research but also provided robust scientific evidence utilizing a deep learning approach to represent neural data for better prediction of online consumer behaviors.