Exploring the Effect of Age and Sex on Subject-Independent EEG-Based Emotion Recognition Methods

In the field of emotion recognition, there are two approaches used to develop predictive models that recognize emotions from EEG signals: subject-dependent and subject-independent. Subject-independent models, although more practical, tend to yield a lower performance due to the high variability of EEG signals between individuals. Recent studies have indicated that incorporating prior demographic information about individuals can improve the accuracy of subject-independent approaches. Some studies have supported this claim by showing that including individuals’ sex can boost model accuracy. However, until now, no one has used interpretable models to measure to what extent demographics can enhance subject-independent approaches. In this work, we follow this direction by using a logistic regression model to correlate the output of a deep learning model with subjects’ age and sex, thereby evaluating whether these factors impact emotion prediction. Our analysis indicates that the ‘sex’ variable significantly influenced the predictions of the deep learning model in three out of five emotions, whereas ‘age’ does not have any effect. These findings suggest that sex is a factor that needs to be considered when designing EEG-based emotion recognition models, which could lead to more robust subject-independent models with potential applications in areas such as healthcare, education, and marketing.