Juvencio Santos Nobre
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Juvencio Santos Nobre has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
- Julio Motta Singer
Departamento de Estatística, Universidade de São Paulo, São Paulo, Brazil
6 shared publications - Caio Lucidius Naberezny Azevedo
IME -Universidade Estadual de Campinas São Paulo Brazil
4 shared publications - Luis Gustavo Bastos Pinho
Universidade Federal do Ceará – Fortaleza (CE), Brazil
3 shared publications - Regis C. Pinheiro Marques
Lightkurve curva de luz exoplanetas periodograma
3 shared publications Gauss Moutinho Cordeiro
2 shared publications- João Victor B. de Freitas
Departamento de Estatística, Instituto de Matemática, Estatística e Computação Científica, Universidade Estadual de Campinas, Campinas, Brazil
2 shared publications
Latest Publications
- Open Access
- 1 December 2025
- Published by Universidad Nacional de Colombia in Revista Colombiana de Estadística
- Vol. 48 (3) , 397-431
A Bayesian Approach for Unit-Gamma Regression through Reparameterization in the Mean and Quantiles
In recent years, considerable attention has been devoted to statistical models designed for continuous doubly bounded dependent variables, defined on a finite interval (a, b) with known limits satisfying −∞ < a < b < ∞. A predominant focus, however, has been on the special case in which the response is restricted to the standard unit interval, namely (0, 1). In this context, we introduce a new class of quantile regression models based on the unit-gamma distribution, along with a mean-based model. For inference, we adopt a Bayesian framework, employing Markov Chain Monte Carlo (MCMC) stochastic simulation algorithms. We conducted extensive simulations to verify the computational implementation and assess the properties of the Bayesian estimators. All analyses were carried out using the nimble package in R, which provides a flexible environment for building and fitting Bayesian hierarchical models. Our results suggest that the proposed Bayesian approach provides unbiased and consistent estimators for all model parameters. Additionally, we perform a detailed model fit assessment, comparing the proposed models with several established alternatives, and conduct an influence analysis to identify potential outliers or influential data points. Finally, we apply the proposed models to a real-world dataset, demonstrating their practical utility and systematically comparing their performance with that of existing models commonly used in the literature.
- 1 December 2025
- Published by Universidad Nacional de Colombia in Revista Colombiana de Estadística
- Vol. 48 (3) , 397-431
Open AccessA Bayesian Approach for Unit-Gamma Regression through Reparameterization in the Mean and Quantiles
In recent years, considerable attention has been devoted to statistical models designed for continuous doubly bounded dependent variables, defined on a finite interval (a, b) with known limits satisfying −∞ < a < b < ∞. A predominant focus, however, has been on the special case in which the response is restricted to the standard unit interval, namely (0, 1). In this context, we introduce a new class of quantile regression models based on the unit-gamma distribution, along with a mean-based model. For inference, we adopt a Bayesian framework, employing Markov Chain Monte Carlo (MCMC) stochastic simulation algorithms. We conducted extensive simulations to verify the computational implementation and assess the properties of the Bayesian estimators. All analyses were carried out using the nimble package in R, which provides a flexible environment for building and fitting Bayesian hierarchical models. Our results suggest that the proposed Bayesian approach provides unbiased and consistent estimators for all model parameters. Additionally, we perform a detailed model fit assessment, comparing the proposed models with several established alternatives, and conduct an influence analysis to identify potential outliers or influential data points. Finally, we apply the proposed models to a real-world dataset, demonstrating their practical utility and systematically comparing their performance with that of existing models commonly used in the literature.
- Open Access
- 23 September 2025
- Published by Instituto Federal de Educacao, Ciencia e Tecnologia Sul-Rio-Grandense in Revista Thema
- Vol. 24 (2) , 1-17
DESENVOLVIMENTO DE MODELOS PARA AVALIAR O RISCO RELATIVO INSTANTÂNEO EM PACIENTES DE COVID-19 EM ESTÁGIOS INICIAIS E TARDIOS PARA USO COMO UM SISTEMA DE ESCORE
COVID-19, recognized as an outbreak in January 2020 and declared a pandemic shortly after, remains a clinical concern. This study developed a predictive model to classify patients by Relative Risk, which can be employed as a scoring system. It is a retrospective cohort study using data from patients admitted to the National Institute for Infectious Diseases "Lazzaro Spallanzani" (Italy) between January 29 and March 28, 2020. Cox Proportional Hazards models were created for patient admission (n = 218) and for the late stage of the disease (n = 385). The models were validated through Machine Learning and tested for violations of Cox assumptions and statistical power. Coefficients were extracted, and equations were derived to calculate the Relative Risk. For Early Stage Relative Risk (ERR), the following variables were used: Age Hazard Ratio (HR): 1.05 [95% CI: 1.02 - 1.08], RDW-SD (HR): 1.08 [95% CI, 1.02 - 1.14], Lymphopenia (HR): 3.23 [95% CI, 1.32 - 7.95], and Obesity (HR): 2.88 [95% CI, 1.08 - 7.69]. For Late Stage Relative Risk (LRR), the variables were: LDH (HR): 7.41 [95% CI, 3.00 - 18.28] and RDW-CV (HR): 5.74 [95% CI, 2.28 - 14.45]. This study provides models and equations that can be implemented as a scoring system to assist in decision-making by calculating the instantaneous Relative Risk (RR).
- 23 September 2025
- Published by Instituto Federal de Educacao, Ciencia e Tecnologia Sul-Rio-Grandense in Revista Thema
- Vol. 24 (2) , 1-17
Open AccessDESENVOLVIMENTO DE MODELOS PARA AVALIAR O RISCO RELATIVO INSTANTÂNEO EM PACIENTES DE COVID-19 EM ESTÁGIOS INICIAIS E TARDIOS PARA USO COMO UM SISTEMA DE ESCORE
COVID-19, recognized as an outbreak in January 2020 and declared a pandemic shortly after, remains a clinical concern. This study developed a predictive model to classify patients by Relative Risk, which can be employed as a scoring system. It is a retrospective cohort study using data from patients admitted to the National Institute for Infectious Diseases "Lazzaro Spallanzani" (Italy) between January 29 and March 28, 2020. Cox Proportional Hazards models were created for patient admission (n = 218) and for the late stage of the disease (n = 385). The models were validated through Machine Learning and tested for violations of Cox assumptions and statistical power. Coefficients were extracted, and equations were derived to calculate the Relative Risk. For Early Stage Relative Risk (ERR), the following variables were used: Age Hazard Ratio (HR): 1.05 [95% CI: 1.02 - 1.08], RDW-SD (HR): 1.08 [95% CI, 1.02 - 1.14], Lymphopenia (HR): 3.23 [95% CI, 1.32 - 7.95], and Obesity (HR): 2.88 [95% CI, 1.08 - 7.69]. For Late Stage Relative Risk (LRR), the variables were: LDH (HR): 7.41 [95% CI, 3.00 - 18.28] and RDW-CV (HR): 5.74 [95% CI, 2.28 - 14.45]. This study provides models and equations that can be implemented as a scoring system to assist in decision-making by calculating the instantaneous Relative Risk (RR).
- 25 July 2025
- Published by Elsevier in Signal Processing
- Vol. 238
Alternative Cholesky Decomposition and family of scale mixture of Normal distribution: A joint modeling approach
- 25 July 2025
- Published by Elsevier in Signal Processing
- Vol. 238
Alternative Cholesky Decomposition and family of scale mixture of Normal distribution: A joint modeling approach
- Open Access
- 6 January 2025
- Published by Brazilian Journals in Caderno Pedagógico
- Vol. 22 (1) , e13290
Análise de confiabilidade como metodologia para gestão da durabilidade de medidores de água
Os medidores de água são instrumentos importantes para a gestão dos recursos hídricos e na garantia de uma cobrança justa para as companhias de saneamento e seus usuários. Normas e legislações específicas regulamentam a construção, verificação e conformidade desses equipamentos, exigindo que as companhias mantenham os medidores em conformidade para assegurar medições precisas. Contudo, a maioria dos estudos na área foca-se apenas em perdas e submedições, negligenciando a análise da conformidade e a manutenção das características metrológicas ao longo do tempo. O presente estudo propõe avaliar a durabilidade dos medidores de água, considerando os fatores relacionados ao fabricante e aos processos de recuperação, por meio de técnicas de análise de sobrevivência/confiabilidade. Dados de verificações intermediárias de uma companhia de saneamento foram analisados por meio do estimador de Kaplan-Meier, e fatores de influência, como fabricante e reparação dos hidrômetros, foram avaliados através da utilização do teste log-rank. Os resultados permitiram estimar a confiabilidade dos medidores para a política vigente da empresa e sugerir o uso dos dados de confiabilidade para orientar estratégias de substituição de hidrômetros.
- 6 January 2025
- Published by Brazilian Journals in Caderno Pedagógico
- Vol. 22 (1) , e13290
Open AccessAnálise de confiabilidade como metodologia para gestão da durabilidade de medidores de água
Os medidores de água são instrumentos importantes para a gestão dos recursos hídricos e na garantia de uma cobrança justa para as companhias de saneamento e seus usuários. Normas e legislações específicas regulamentam a construção, verificação e conformidade desses equipamentos, exigindo que as companhias mantenham os medidores em conformidade para assegurar medições precisas. Contudo, a maioria dos estudos na área foca-se apenas em perdas e submedições, negligenciando a análise da conformidade e a manutenção das características metrológicas ao longo do tempo. O presente estudo propõe avaliar a durabilidade dos medidores de água, considerando os fatores relacionados ao fabricante e aos processos de recuperação, por meio de técnicas de análise de sobrevivência/confiabilidade. Dados de verificações intermediárias de uma companhia de saneamento foram analisados por meio do estimador de Kaplan-Meier, e fatores de influência, como fabricante e reparação dos hidrômetros, foram avaliados através da utilização do teste log-rank. Os resultados permitiram estimar a confiabilidade dos medidores para a política vigente da empresa e sugerir o uso dos dados de confiabilidade para orientar estratégias de substituição de hidrômetros.
- Open Access
- 1 January 2025
- Published by FapUNIFESP (SciELO) in Anais da Academia Brasileira de Ciências
- Vol. 97 (4) , e20250071
Bayesian inference for zero-and/or-one augmentedunit-gamma
In this paper, we propose a new distribution based on a mixture of the unit-gamma distribution with a degenerate distribution at (c) (0 or1) or with a Bernoulli distribution. This novel approach is particularlyuseful for addressing excess zeros and/or ones in the data with limitedsupport in ( 0 , 1 ).Our approach considers Bayesian parameter estimation, residual, and influence analysis, as well as model comparison methods. We illustratethe theory by applying it to a real-world problem, with all posteriorquantities obtained using Markov Chain Monte Carlo (MCMC) methods.
- 1 January 2025
- Published by FapUNIFESP (SciELO) in Anais da Academia Brasileira de Ciências
- Vol. 97 (4) , e20250071
Open AccessBayesian inference for zero-and/or-one augmentedunit-gamma
In this paper, we propose a new distribution based on a mixture of the unit-gamma distribution with a degenerate distribution at (c) (0 or1) or with a Bernoulli distribution. This novel approach is particularlyuseful for addressing excess zeros and/or ones in the data with limitedsupport in ( 0 , 1 ).Our approach considers Bayesian parameter estimation, residual, and influence analysis, as well as model comparison methods. We illustratethe theory by applying it to a real-world problem, with all posteriorquantities obtained using Markov Chain Monte Carlo (MCMC) methods.
- Open Access
- 2 September 2024
- Published by Brazilian Journals in Caderno Pedagógico
- Vol. 21 (9) , e7690
Bayesian inference for unit Gamma distribution
In this article, we develop Bayesian inference for the unit-Gamma (GU) distribution, initially introduced by Grassia (1977). This distribution is highly flexible, allowing it to take various forms within the interval (0,1), encompassing both symmetric and asymmetric shapes. Such flexibility makes it an attractive alternative to traditional distributions in this range, like the Kumaraswamy and beta models. We propose a parameterization based on quantiles, a particularly advantageous approach when dealing with datasets that include outliers, since the median, for instance, is a more robust estimator compared to the mean in such cases. Our work covers parameter estimation, model fit assessment, model comparison, and influence analysis. All procedures were implemented using Markov Chain Monte Carlo (MCMC) methods via Just Another Gibbs Sampling (JAGS) through the R2jags package in R, an open-source software. Furthermore, we demonstrate the effectiveness of this methodology by applying it to a real-world dataset, highlighting its practical utility.
- 2 September 2024
- Published by Brazilian Journals in Caderno Pedagógico
- Vol. 21 (9) , e7690
Open AccessBayesian inference for unit Gamma distribution
In this article, we develop Bayesian inference for the unit-Gamma (GU) distribution, initially introduced by Grassia (1977). This distribution is highly flexible, allowing it to take various forms within the interval (0,1), encompassing both symmetric and asymmetric shapes. Such flexibility makes it an attractive alternative to traditional distributions in this range, like the Kumaraswamy and beta models. We propose a parameterization based on quantiles, a particularly advantageous approach when dealing with datasets that include outliers, since the median, for instance, is a more robust estimator compared to the mean in such cases. Our work covers parameter estimation, model fit assessment, model comparison, and influence analysis. All procedures were implemented using Markov Chain Monte Carlo (MCMC) methods via Just Another Gibbs Sampling (JAGS) through the R2jags package in R, an open-source software. Furthermore, we demonstrate the effectiveness of this methodology by applying it to a real-world dataset, highlighting its practical utility.
- 10 February 2024
- Published by Springer Nature in Statistics and Computing
- Vol. 34 (2) , 1-18
A robust quantile regression for bounded variables based on the Kumaraswamy Rectangular distribution
- 10 February 2024
- Published by Springer Nature in Statistics and Computing
- Vol. 34 (2) , 1-18
A robust quantile regression for bounded variables based on the Kumaraswamy Rectangular distribution
- 1 January 2024
- Published by International Press of Boston in Statistics and Its Interface
- Vol. 17 (4) , 715-731
Flexible quasi-beta prime regression models for dependent continuous positive data
International Press of Boston - publishers of scholarly mathematical and scientific journals and books
- 1 January 2024
- Published by International Press of Boston in Statistics and Its Interface
- Vol. 17 (4) , 715-731
Flexible quasi-beta prime regression models for dependent continuous positive data
International Press of Boston - publishers of scholarly mathematical and scientific journals and books
- Open Access
- 1 January 2022
- Published by FapUNIFESP (SciELO) in Anais da Academia Brasileira de Ciências
- Vol. 94 (3) , e20201542
Continuous Probability Distributions generated by the PIPE Algorithm
We investigate the use of the Probabilistic Incremental Programming Evolution (PIPE) algorithm as a tool to construct continuous cumulative distribution functions to model given data sets. The PIPE algorithm can generate several candidate functions to fit the empirical distribution of data. These candidates are generated by following a set of probability rules. The set of rules is then evolved over a number of iterations to generate better candidates regarding some optimality criteria. This approach rivals that of generated distribution, obtained by adding parameters to existing probability distributions. There are two main advantages for this method. The first is that it is possible to explicitly control the complexity of the candidate functions, by specifying which mathematical functions and operators can be used and how lengthy the mathematical expression of the candidate can be. The second advantage is that this approach deals with model selection and estimation at the same time. The overall performance in both simulated and real data was very satisfying. For the real data applications, the PIPE algorithm obtained better likelihoods for the data when compared to existing models, but with remarkably simpler mathematical expressions.
- 1 January 2022
- Published by FapUNIFESP (SciELO) in Anais da Academia Brasileira de Ciências
- Vol. 94 (3) , e20201542
Open AccessContinuous Probability Distributions generated by the PIPE Algorithm
We investigate the use of the Probabilistic Incremental Programming Evolution (PIPE) algorithm as a tool to construct continuous cumulative distribution functions to model given data sets. The PIPE algorithm can generate several candidate functions to fit the empirical distribution of data. These candidates are generated by following a set of probability rules. The set of rules is then evolved over a number of iterations to generate better candidates regarding some optimality criteria. This approach rivals that of generated distribution, obtained by adding parameters to existing probability distributions. There are two main advantages for this method. The first is that it is possible to explicitly control the complexity of the candidate functions, by specifying which mathematical functions and operators can be used and how lengthy the mathematical expression of the candidate can be. The second advantage is that this approach deals with model selection and estimation at the same time. The overall performance in both simulated and real data was very satisfying. For the real data applications, the PIPE algorithm obtained better likelihoods for the data when compared to existing models, but with remarkably simpler mathematical expressions.
- 11 August 2021
- Published by Taylor & Francis in Journal of Applied Statistics
- Vol. 49 (15) , 3784-3803
A new approach to modeling positive random variables with repeated measures
In many situations, it is common to have more than one observation per experimental unit, thus generating the experiments with repeated measures. In the modeling of such experiments, it is necessary to consider and model the intra-unit dependency structure. In the literature, there are several proposals to model positive continuous data with repeated measures. In this paper, we propose one more with the generalization of the beta prime regression model. We consider the possibility of dependence between observations of the same unit. Residuals and diagnostic tools also are discussed. To evaluate the finite-sample performance of the estimators, using different correlation matrices and distributions, we conducted a Monte Carlo simulation study. The methodology proposed is illustrated with an analysis of a real data set. Finally, we create an package for easy access to publicly available the methodology described in this paper.- 11 August 2021
- Published by Taylor & Francis in Journal of Applied Statistics
- Vol. 49 (15) , 3784-3803
A new approach to modeling positive random variables with repeated measures
In many situations, it is common to have more than one observation per experimental unit, thus generating the experiments with repeated measures. In the modeling of such experiments, it is necessary to consider and model the intra-unit dependency structure. In the literature, there are several proposals to model positive continuous data with repeated measures. In this paper, we propose one more with the generalization of the beta prime regression model. We consider the possibility of dependence between observations of the same unit. Residuals and diagnostic tools also are discussed. To evaluate the finite-sample performance of the estimators, using different correlation matrices and distributions, we conducted a Monte Carlo simulation study. The methodology proposed is illustrated with an analysis of a real data set. Finally, we create an package for easy access to publicly available the methodology described in this paper.