Sparse Coding Techniques with Deep Neural Network for EEG Auto Encoding: Enhancing Feature Representation Efficiency

Research on emotion classification using Brain-Computer Interface (BCI) systems is an intriguing aspect. Deep learning has recently been used to classify emotions in BCI systems, and better results have been seen when compared to more conventional classification techniques. Although the high dimensi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International Conference on Recent Advances in Information Technology (Online) S. 1 - 8
Hauptverfasser: Rama, M., Nalini, T, Khyathi, P, Saikiran, N., Silpa, Ch
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 06.03.2025
Schlagworte:
Autoencoders
Autoencoding
Deep learning
Electroencephalography
Encoding
Feature Learning
Feature Representation
Noise
Representation learning
Root mean square
Scalability
Signal analysis
Signal reconstruction
SparseCoding
ISSN:2994-287X
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Research on emotion classification using Brain-Computer Interface (BCI) systems is an intriguing aspect. Deep learning has recently been used to classify emotions in BCI systems, and better results have been seen when compared to more conventional classification techniques. Although the high dimensionality and noise of Electroencephalography (EEG) signals make it difficult to research brain activity is essential for efficient feature representation. To improve the effectiveness of feature representation, describe a unique method in this paper for EEG autoencoding that combines Sparse Coding Techniques with Deep Neural Networks (SCT-DNN).. Using several encoding and decoding layers, the deep auto-encoder is intended to learn hierarchical representations that capture crucial aspects of the EEG signal. Promote efficient representation and reconstruction of EEG signals by enticing the machine to learn sparse and discriminative features through the use of sparse coding throughout the encoding phase. Results from the experiment demonstrate the advantages of SCT-DNN for learning EEG features and provide insights into the underlying brain processes. Through the synergistic use of deep learning and sparse coding, this proposed work optimizes the efficiency of feature representation, furthering the field of EEG signal analysis.
ISSN:2994-287X
DOI:10.1109/RAIT65068.2025.11088927