Learning from imbalanced data sets with weighted cross-entropy function

Página do item simplificado
dc.creatorYuri Sousa Aurelio
dc.creatorGustavo Matheus de Almeida
dc.creatorCristiano Leite de Castro
dc.creatorAntonio de Padua Braga
dc.date.accessioned2025-05-15T14:04:14Z
dc.date.accessioned2025-09-09T00:03:35Z
dc.date.available2025-05-15T14:04:14Z
dc.date.issued2019
dc.identifier.doihttps://doi.org/10.1007/s11063-018-09977-1
dc.identifier.issn1370-4621
dc.identifier.urihttps://hdl.handle.net/1843/82299
dc.languageeng
dc.publisherUniversidade Federal de Minas Gerais
dc.relation.ispartofNeural processing letters
dc.rightsAcesso Restrito
dc.subjectRedes neurais (Computação)
dc.subject.otherThe number of samples commonly differs from one class to another in classification problems. This problem, known as the imbalanced data set problem, arises in most real-world applications. The point is that most current inductive learning principles resides on a sum of squared errors that do not take priors into account, which generally results in a classification bias towards the majority class.
dc.titleLearning from imbalanced data sets with weighted cross-entropy function
dc.typeArtigo de periódico
local.citation.epage1949
local.citation.spage1937
local.citation.volume50
local.description.resumoThis paper presents a novel approach to deal with the imbalanced data set problem in neural networks by incorporating prior probabilities into a cost-sensitive cross-entropy error function. Several classical benchmarks were tested for performance evaluation using different metrics, namely G-Mean, area under the ROC curve (AUC), adjusted G-Mean, Accuracy, True Positive Rate, True Negative Rate and F1-score. The obtained results were compared to well-known algorithms and showed the effectiveness and robustness of the proposed approach, which results in well-balanced classifiers given different imbalance scenarios.
local.publisher.countryBrasil
local.publisher.departmentENG - DEPARTAMENTO DE ENGENHARIA ELÉTRICA
local.publisher.departmentENG - DEPARTAMENTO DE ENGENHARIA ELETRÔNICA
local.publisher.departmentENG - DEPARTAMENTO DE ENGENHARIA QUÍMICA
local.publisher.initialsUFMG
local.url.externahttps://link.springer.com/article/10.1007/s11063-018-09977-1

Arquivos

Licença do pacote

Agora exibindo 1 - 1 de 1
Imagem de Miniatura
Nome:
License.txt
Tamanho:
1.99 KB
Formato:
Plain Text
Descrição:
Baixar

Coleções

Artigo de Periódico

Logo UFMG