Enhancing EEG-Based Emotion Detection with Hybrid Models: Insights from DEAP Dataset Applications

Emotion detection using EEG signals is a rapidly growing field with significant potential in mental health diagnostics and human-computer interaction. This study leverages the DEAP dataset to explore and evaluate various machine learning and deep learning techniques for emotion recognition. Extensive experimentation was conducted with models such as K-Nearest Neighbors (KNN), Support Vector Machines (SVM), Decision Trees (DT), and Random Forests (RF), Bidirectional Long Short-Term Memory (BiLSTM), Gated Recurrent Units (GRU), Convolutional Neural Networks (CNN), Autoencoders, and Transformers, achieving an accuracy of up to 90%. These results demonstrate the effectiveness of advanced neural architectures in decoding emotional states from EEG signals. Furthermore, SHapley Additive exPlanations (SHAP) were employed to interpret the model predictions, enhancing transparency and providing deeper insights into the contribution of individual features to the decision-making process. In addition, A real-time emotion detection system was developed, enabling instantaneous classification of emotional states from EEG signals, which has significant implications for real-world applications such as affective computing, neurofeedback, and human-computer interaction. This real-time capability, combined with the high accuracy and interpretability of the models, ensures that the decisions are not only accurate but also understandable, fostering trust and enabling more informed application in critical areas such as mental health and human-computer interaction.