EEG-Based Emotion Detection Using Roberts Similarity and PSO Feature Selection

The detection of emotions through electroencephalography (EEG) signals has gained significant attention due to its potential applications in areas such as mental health diagnosis, human-computer interaction, and affective computing. This study proposes a novel emotion detection framework that integrates EEG-based signal processing with advanced machine learning and deep learning techniques. To enhance feature extraction, an Extra Trees Classifier is utilized to identify the most relevant features from EEG signals, which are then optimized using the PSOFeatureModelSelector. The feature selection model employs a combination of K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Random Forest (RF) to refine the selection process and improve classification accuracy.

For the classification task, the study leverages both traditional machine learning methods and deep learning architectures. A hybrid CNN-LSTM model is utilized, which effectively captures spatial and temporal dependencies in EEG data. Additionally, a Roberta Similarity-based approach is integrated to quantify the similarity between EEG signal patterns, improving the overall emotion classification process. The framework also explores the use of DeepConvNet, a convolutional neural network-based model, for extracting hierarchical features from raw EEG data, enhancing the detection of complex emotional states.

The results demonstrate the efficacy of combining traditional machine learning models with cutting-edge deep learning architectures, highlighting improvements in accuracy, robustness, and real-time processing capabilities for emotion detection from EEG signals. This work contributes to advancing emotion-aware systems that can interact with users in a more personalized and empathetic manner, paving the way for future applications in healthcare, entertainment, and human-machine interaction.

Keywords: EEG, Emotion Detection, Extra Trees Classifier, Feature Extraction, PSOFeatureModelSelector, KNN, SVM, Random Forest, CNN-LSTM, Roberta Similarity, DeepConvNet, Machine Learning, Deep Learning, Signal Processing, Affective Computing, Human-Computer Interaction, Mental Health, Real-time Processing.