A fair performance comparison between complex-valued and real-valued neural networks for disease detection

Vista sencilla de ítem
dc.contributor.authorJojoa Acosta, Mario Fernando
dc.contributor.authorGarcía-Zapirain, Begoña
dc.contributor.authorPercybrooks, Winston S.
dc.date.accessioned2025-07-18T12:04:24Z
dc.date.available2025-07-18T12:04:24Z
dc.date.issued2022-08-04
dc.date.updated2025-07-18T12:04:24Z
dc.description.abstractOur aim is to contribute to the classification of anomalous patterns in biosignals using this novel approach. We specifically focus on melanoma and heart murmurs. We use a comparative study of two convolution networks in the Complex and Real numerical domains. The idea is to obtain a powerful approach for building portable systems for early disease detection. Two similar algorithmic structures were chosen so that there is no bias determined by the number of parameters to train. Three clinical data sets, ISIC2017, PH2, and Pascal, were used to carry out the experiments. Mean comparison hypothesis tests were performed to ensure statistical objectivity in the conclusions. In all cases, complex-valued networks presented a superior performance for the Precision, Recall, F1 Score, Accuracy, and Specificity metrics in the detection of associated anomalies. The best complex number-based classifier obtained in the Receiving Operating Characteristic (ROC) space presents a Euclidean distance of 0.26127 with respect to the ideal classifier, as opposed to the best real number-based classifier, whose Euclidean distance to the ideal is 0.36022 for the same task of melanoma detection. The 27.46% superiority in this metric, as in the others reported in this work, suggests that complex-valued networks have a greater ability to extract features for more efficient discrimination in the dataset.en
dc.identifier.citationJojoa, M., Garcia-Zapirain, B., & Percybrooks, W. (2022). A fair performance comparison between complex-valued and real-valued neural networks for disease detection. Diagnostics, 12(8). https://doi.org/10.3390/DIAGNOSTICS12081893
dc.identifier.doi10.3390/DIAGNOSTICS12081893
dc.identifier.eissn2075-4418
dc.identifier.urihttps://hdl.handle.net/20.500.14454/3252
dc.language.isoeng
dc.publisherMDPI
dc.rights© 2022 by the authors
dc.subject.otherComplex numbers
dc.subject.otherComplex-valued convolution neural networks
dc.subject.otherComplex-valued deep learning
dc.subject.otherFair performance comparison
dc.subject.otherReal-valued neural networks
dc.titleA fair performance comparison between complex-valued and real-valued neural networks for disease detectionen
dc.typejournal article
dcterms.accessRightsopen access
oaire.citation.issue8
oaire.citation.titleDiagnostics
oaire.citation.volume12
oaire.licenseConditionhttps://creativecommons.org/licenses/by/4.0/
oaire.versionVoR
Archivos
Bloque original
Mostrando 1 - 1 de 1
Miniatura
Nombre:
jojoa_fair_2022.pdf
Tamaño:
2.22 MB
Formato:
Adobe Portable Document Format
Descargar
Colecciones
Artículos