NMNet: Spatial-Temporal Transformer for EEG Signal Analysis in Neuromarketing

In this work, we propose a deep neural network, NMNet, to predict consumer preferences for E-commerce products by analyzing neural activity captured through EEG signals. Our approach utilizes a transformer-based model to extract features from EEG signals in both the temporal and spatial domains. We evaluate our approach using a dataset consisting of 1050 EEG signals collected from 25 participants. We compare our approach against the existing baseline and consistently outperform across all evaluation metrics. In order to assess the generalization ability of our proposed method, we employ a diverse set of stimuli during the evaluation phase. This enables us to thoroughly evaluate the performance of our method across a range of different stimuli, providing valuable insights into its effectiveness and adaptability.