ABSTRACT
Creating robust and explainable statistical learning models is essential in credit risk management. For this purpose, equally spaced or frequent discretization is the de facto choice when building predictive models. The methods above have limitations, given that when the discretization procedure is constrained, the underlying patterns are lost. This study introduces an innovative approach by combining traditional discretization techniques with clustering-based discretization, specifically k means and Gaussian mixture models. The study proposes two combinations: Discrete Competitive Combination (DCC) and Discrete Exhaustive Combination (DEC). Discrete Competitive Combination selects features based on the discretization method that performs better on each feature, whereas Discrete Exhaustive Combination includes every discretization method to complement the information not captured by each technique. The proposed combinations were tested on 11 different credit risk datasets by fitting a logistic regression model using the weight of evidence transformation over the training partition and contrasted over the validation partition. The experimental findings showed that both combinations similarly outperform individual methods for the logistic regression without compromising the computational efficiency. More importantly, the proposed method is a feasible and competitive alternative to conventional methods without reducing explainability.
Subject(s)
Learning , Models, Statistical , Logistic Models , Cluster AnalysisABSTRACT
(1) Background: Carbohydrates are the most important source of nutritional energy for the human body. Carbohydrate digestion, metabolism, and their role in the gut microbiota modulation are the focus of multiple studies. The objective of this weight of evidence systematic review is to investigate the potential relationship between ingested carbohydrates and the gut microbiota composition at different taxonomic levels. (2) Methods: Weight of evidence and information value techniques were used to evaluate the relationship between dietary carbohydrates and the relative abundance of different bacterial taxa in the gut microbiota. (3) Results: The obtained results show that the types of carbohydrates that have a high information value are: soluble fiber with Bacteroides increase, insoluble fiber with Bacteroides and Actinobacteria increase, and Firmicutes decrease. Oligosaccharides with Lactobacillus increase and Enterococcus decrease. Gelatinized starches with Prevotella increase. Starches and resistant starches with Blautia decrease and Firmicutes increase. (4) Conclusions: This work provides, for the first time, an integrative review of the subject by using statistical techniques that have not been previously employed in microbiota reviews.
ABSTRACT
Se realiza análisis prospectivo aleatorio a 198 pacientes con meningoencefalitis aséptica (MEA) en el servicio de medicina del hospital provincial "Dr. antonio Luaces Iraola" de Ciego de Avila. Se les efectuo estudios con sueros pareados para determinar anticuerpos neutralizantes a enterovirus, pruebas de hemoaglutinación, neutralización viral para arbovirus, hemolítica para leptospira y estudio virológico de heces fecales buscando enterovirus. En 36.8 porciento de los pacientes se encontró como agente etiológico un enterovirus. El segundo agente en frecuencia fue la leptospira en 14.6 porciento. Se determinó el agente etiológico en 51.4 porciento de los pacientes estudiados. (AU)
Subject(s)
Humans , Meningoencephalitis , CubaABSTRACT
Se realiza análisis prospectivo aleatorio a 198 pacientes con meningoencefalitis aséptica (MEA) ingresados en el servicio de medicina del hospital provincial ®Dr. Antonio Luaces Iraola¼ de Ciego de Avila, Cuba. Se les efectuó estudios con sueros pareados para determinar anticuerpos neutralizantes a enterovirus, pruebas de hemoglutinación, neutralización viral para arbovirus, hemolítica para leptospira y estudio virológico de heces fecales buscando enterovirus. En 36.8 por ciento de los pacientes se encontró como agente etiológico un enterovirus. El segundo agente en frecuencia fue la leptospira en 14.6 por ciento. Se determinó el agente etiológico en 51.4 por ciento de los pacientes estudiados
