Genetics-Based Machine Learning for Rule Induction: State of the Art, Taxonomy, and Comparative Study

Fernández, Alberto; García, Salvador; Luengo, Julián; Bernadó-Mansilla, Ester; Herrera, F.

doi:10.1109/tevc.2009.2039140

Cited by 144 publications

(85 citation statements)

References 71 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In section 4, we will assess the performance of our approach. At first we will compare our results to those gathered by Fernandez et al with 22 state-of-the-art classifiers in the context of imbalanced data [1], showing our approach can be applied on more general datasets. Secondly, we will compare our approach to C4.5 -a state-of-the-art decision tree algorithm -and C4.5-CS -an adaptation of the C4.5 algorithm to imbalanced data -on real hospital data.…”

Section: Introductionmentioning

confidence: 92%

“…Fernandez et al performed a comparison of 22 classification rule mining algorithms on imbalanced datasets and provided material to compare to their results [1]. Since our algorithm is designed to handle discrete attributes, datasets with less continuous attributes were preferred.…”

Section: Experiments On Imbalanced Benchmarks Datasetsmentioning

confidence: 99%

“…Additionally, more and more information is available and collected nowadays: algorithms must be able to deal with larger datasets. According to Fernández et al dealing with large datasets is still a challenge that needs to be addressed [1]. This work is a part of the OPCYCLIN project -an industrial project involving Alicante company, hospitals and academics as partners -that aims at providing a tool to optimize screening of patients for clinical trials.…”

Section: Introductionmentioning

confidence: 99%

“…To avoid this problem we will focus on classification rule mining approaches. The majority of state-of-the-art classification algorithms will have trouble to deal with imbalanced data because they use Accuracy to build their predictive model [1]. Accuracy focuses on counting good classifications obtained by a given algorithm: true positives and true negatives.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

MOCA-I: Discovering Rules and Guiding Decision Maker in the Context of Partial Classification in Large and Imbalanced Datasets

Jacques

Taillard²,

Delerue³

et al. 2013

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 92%

Section: Experiments On Imbalanced Benchmarks Datasetsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

MOCA-I: Discovering Rules and Guiding Decision Maker in the Context of Partial Classification in Large and Imbalanced Datasets

Jacques

Taillard²,

Delerue³

et al. 2013

Lecture Notes in Computer Science

View full text Add to dashboard Cite

“…Destas bases toda umaárea se desenvolveu, tanto pela busca de novas aplicações fora daárea de controle (reconhecimento de padrões, regressão, entre outras) e de novas meta-heurísticas (por exemplo, Programação Genérica -PG), quanto ao propor outros mecanismos para avaliação, sintetização e formação de um SIF [48,78]. Toda uma literatura de SFG foi eé constantemente elaborada, tais como livros [49,86,113], edições especiais em revistas [7,8,162] e diversos exames sobre aárea [48,76,78,98].…”

Section: Introductionunclassified

Gpfis: Um Sistema Fuzzy-Genético Genérico Baseado Em Programação Genética

KOSHIYAMA¹

View full text Add to dashboard Cite

AgradecimentosAo povo brasileiro, ao CNPq e a PUC-Rio pelos auxílios e estrutura concedida, sem os quais este trabalho não poderia ter sido realizado.A minha família, em destaque meus irmãos -Wladmir e Kallyah -e pais -Adriano e Sayonara -por todos os momentos abdicados da minha presença, devido a este manuscrito e demais projetos. Agradeço e saibam que temos uma ligação de afeto para além do infinito.A minha orientadora, Professora Marley Vellasco, que sempre me estimulou a produzir cientificamente, me incentivando nas inúmeras reuniões e revisões de texto, além da oportunidade de trabalhar em projetos. Também ao meu co-orientador, Professor Ricardo Tanscheit, pelas excelentes conversas e histórias compartilhadas e por ter me permitido absorver um pouco do seu método "minimalista". A ambos o meu maior respeito e admiração.Aos Também a pessoas com quem pude mais aprender do que transmitir, e com as quais formou-se uma amizade sólida. Iury Steiner, João Zanni, Pedro Milet e Douglas Dias, a vocês obrigado por todas as risadas e experiência passada. Ainda, a Tatiana Escovedo e Rubens Nascimento que me adotaram como filho e partilharam comigo trabalhos, pizzas e viagens.Aos professores Marco Pacheco, Wagner Tassinari,Álvaro Veiga, Carlos Tomei e Maria Lorenzon, dos quais pude absorver, cada um a sua maneira, a forma de lidar e transmitir a ciência, sempre com rigor e crítica, e me permitiram também vislumbrar a aplicação desta fora dos muros da universidade. In memorian a Padre Libanio pela amizade partilhada e exemplo de vida.Por fim, ao corpo de funcionários do Departamento de Engenharia Elétrica, notavelmente Alcina, Mara, Márcia, Isnarde e Miriam pelo apoio.

show abstract

Corporate failure prediction using threshold‐based models

Veganzones

2022

Journal of Forecasting

View full text Add to dashboard Cite

Corporate failure prediction literature indicates that models' performance depends on more than the complexity of the prediction method. Recent advances cite the relevance of sampling approaches to model performance.Therefore, this study proposes a novel approach that implements threshold models for corporate failure prediction efforts. Threshold models estimate asymptotically conservative confidence regions, in which the samples are split by size. Then, single-based classifiers and a combination of multiple classifiers are employed in each region to estimate the prediction accuracy. This article offers a comparison of the proposed threshold-based corporate failure model with two benchmark models, previously used in studies in the same context.The empirical results show that the proposed threshold-based model clearly outperforms conventional models, in particular with the combination of multiple classifiers. The superiority of the threshold-based model stems from its ability to discern failed firms, which represent the most important class in financial terms. This study thus provides initial evidence of the utility of threshold models in corporate failure prediction efforts.

show abstract

Genetics-Based Machine Learning for Rule Induction: State of the Art, Taxonomy, and Comparative Study

Cited by 144 publications

References 71 publications

MOCA-I: Discovering Rules and Guiding Decision Maker in the Context of Partial Classification in Large and Imbalanced Datasets

MOCA-I: Discovering Rules and Guiding Decision Maker in the Context of Partial Classification in Large and Imbalanced Datasets

Gpfis: Um Sistema Fuzzy-Genético Genérico Baseado Em Programação Genética

Corporate failure prediction using threshold‐based models

Contact Info

Product

Resources

About