Authors:
Jiří Vaněk
and
Roman Mouček
Affiliation:
University of West Bohemia, Czech Republic
Keyword(s):
Deep Learning, Neural Networks, Stacked Autoencoder, Deep Belief Networks, Classification, Event-related Potentials, P300 Component.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Biomedical Engineering
;
Business Analytics
;
Cardiovascular Technologies
;
Computing and Telecommunications in Cardiology
;
Data Engineering
;
Decision Support Systems
;
Decision Support Systems, Remote Data Analysis
;
Health Engineering and Technology Applications
;
Health Information Systems
;
Knowledge-Based Systems
;
Pattern Recognition and Machine Learning
;
Symbolic Systems
Abstract:
Deep learning techniques have proved to be beneficial in many scientific disciplines and have beaten stateof-
the-art approaches in many applications. The main aim of this article is to improve the success rate of
deep learning algorithms, especially stacked autoencoders, when they are used for detection and classification
of P300 event-related potential component that reflects brain processes related to stimulus evaluation or
categorization. Moreover, the classification results provided by stacked autoencoders are compared with the
classification results given by other classification models and classification results provided by combinations
of various types of neural network layers.