Bayesian Binary Regression Research Articles

Bayesian spatial analysis of incomplete vaccination among children aged 12–23 months in Nigeria

High childhood disease prevalence and under-five mortality rates have been consistently reported in Nigeria. Vaccination is a cost-effective preventive strategy against childhood diseases. Therefore, this study aimed to identify the determinants of Incomplete Vaccination (IV) among children aged 12–23 months in Nigeria. This cross-sectional design study utilized the 2018 Nigeria Demographic and Health Survey (NDHS) dataset. A two-stage cluster sampling technique was used to select women of reproductive age who have children (n = 5475) aged 12–23 months. The outcome variable was IV of children against childhood diseases. Data were analyzed using Integrated Nested Laplace Approximation and Bayesian binary regression models (α0.05). Visualization of incomplete vaccination was produced using the ArcGIS software. Children’s mean age was 15.1 ± 3.2 months and the median number of vaccines received was four. Northern regions contributed largely to the IV. The likelihood of IV was lower among women aged 25–34 years (aOR = 0.67, 95% CI = 0.54–0.82, p < 0.05) and 35–49 years (aOR = 0.59, 95%CI = 0.46–0.77, p < 0.05) compared to younger women in the age group 15–24 years. An increasing level of education reduces the risk of odds of IV. Other predictors of IV were delivery at the health facility (aOR = 0.64, 95% CI = 053–0.76, p < 0.05), and media exposure (aOR = 0.63, 95%CI = 0.54–0.79, p < 0.05). Mothers’ characteristics explained most of the variability in the IV, relatively to smaller overall contributions from the community and state-level factors (p < 0.05). The level of IV against childhood diseases was high in Nigeria. However, disparities exist across the regions and other socioeconomic segments of the population. More efforts are required to improve vaccination sensitization programs and campaigns in Nigeria.

Bayesian Binary Quantile Regression for Modelling Injectable Contraceptive Uptake Among Child Bearing Women in Kenya

Bayesian Binary Quantile Regression for Modelling Injectable Contraceptive Uptake Among Child Bearing Women in Kenya

Bayesian binary regression using power and power reverse link functions: an application to premature birth data

This study aims to determine factors associated (and quantify) with prematurity through binary regression models, considering power and reverse power link functions, with asymmetric characteristics. As criteria for the model selection, the Bayesian Deviance Information Criterion (DIC), predictive evaluation, and residual analysis. All models analyzed presented similar predictive capacity, however, the model with a reverse power logit link function, with asymmetry parameter =0.336 was chosen, since it presented the lowest value of DIC=3,203, residues that indicated a good fitted of the model. There was an association of prematurity with the following variables: maternal - age over 35 years (OR=1.485), with a partner (OR=0.731), and primiparous (OR=1.307); of pregnancy and childbirth - multiple pregnancy (OR=36.360), cesarean childbirth (OR=1.337) and number of prenatal consultations less than seven (OR=3.305); and newborns of white race/skin (OR=0.731) and presence of congenital malformation (OR=2.663). Considering the proposed criteria, an asymmetric link function (reverse power logit) was the most parsimonious for the model. From this, there were high chances of factors associated with the occurrence of prematurity, indicating the need for actions to minimize them.

Exact convergence analysis for metropolis–hastings independence samplers in Wasserstein distances

AbstractUnder mild assumptions, we show that the exact convergence rate in total variation is also exact in weaker Wasserstein distances for the Metropolis–Hastings independence sampler. We develop a new upper and lower bound on the worst-case Wasserstein distance when initialized from points. For an arbitrary point initialization, we show that the convergence rate is the same and matches the convergence rate in total variation. We derive exact convergence expressions for more general Wasserstein distances when initialization is at a specific point. Using optimization, we construct a novel centered independent proposal to develop exact convergence rates in Bayesian quantile regression and many generalized linear model settings. We show that the exact convergence rate can be upper bounded in Bayesian binary response regression (e.g. logistic and probit) when the sample size and dimension grow together.

Modelling the Misuse of Alcohol and Drugs in South Africa Using Bayesian Binary Logistic Regression

Modelling the Misuse of Alcohol and Drugs in South Africa Using Bayesian Binary Logistic Regression

Examining Seat Belt Use Habits in Wyoming: A Bayesian Hierarchical Logit Approach

Wyoming has one of the lowest seat belt law compliance rates in the United States. Understanding the factors associated with seat belt non-use is crucial for the development of appropriate mitigation measures that aid in increasing seat belt usage rates. In this study, Bayesian binary logistic regression models were developed to: i) investigate the factors associated with daytime seat belt non-use using observational survey data; and ii) assess the factors related to seat belt non-use in fatal crashes while accounting for possible intra-class correlation (effects of unobserved factors common to occupants in the same vehicle). The results indicated that males were less likely to use seat belts than females. Also, trips made in pick-up trucks, those made within urban areas, those made during weekdays, and those made by means of vehicles with Wyoming license plates were likely to involve seat belt non-use. Furthermore, counties with thriving oil and gas industries had lower seat belt use rates. It was also inferred that young occupants, rear seat occupants, impairment, weekdays, traveling around midnight, traveling on collector roads, and traveling on local roads were associated with seat belt non-use. A strong correlation in seat belt non-use among occupants within the same vehicle warrants the implementation of the hierarchical logistic model to investigate seat belt use among occupants involved in fatal crashes. The findings of this study highlight remarkable trends in seat belt use habits in Wyoming, which will aid legislatures in proposing road safety measures aimed at raising seat belt usage proportions.

Classification of The Risk of Comorbid Covid-19 Patient at Bengkalis Hospital Using Bayesian Binary Logistics Regression

COVID-19 is an infectious disease caused by SARS-CoV-2 (Severe Acute Respiratory Syndrome Coronavirus 2). This viral infection causes illness with symptoms ranging from mild to severe. The number of deaths from this disease is increasing day by day. A person who is most easily infected with the COVID-19 virus is a person who has a comorbid disease, because the body's immunity decreases due to the impact of a previous illness. The purpose of this study was to determine the comorbid factors that trigger a person's death due to COVID-19. This research uses binary logistic regression with Bayes method parameter estimation. In this study, the predictor variables used were in the form of categories. The results showed that the factors that influence the death of a person on the death of COVID-19 in comorbid diseases are Diabetes Mellitus and Pneumonia.

Investigating the Factors Affecting the Severity of Single-Vehicle Crashes on Urban Roads using Bayesian Binary Probit Regression

Investigating the Factors Affecting the Severity of Single-Vehicle Crashes on Urban Roads using Bayesian Binary Probit Regression

Covid-19 Patient Mortality Risk Classification Using Bayesian Binary Logistic Regression

At the end of 2019 the world was shocked by a new disease caused by SARS-CoV-2 (Severe Acute Respiratory Syndrome Coronavirus 2). The disease is called Covid-19 (Coronavirus Disease). The mortality rate due to disease is increasing every day. In Indonesia as of April 2021, confirmed Covid-19 patients who died reached 42,530 patients, seeing the high mortality rate of Covid-19 patients so it needs to be studied further so that the risk of death of these Covid-19 patients can be minimized. This research utilizing binary logistic regression with Bayesian method parameter estimation. In this study, the predictor variables used were in the form of categories that each category in the predictor variables was assumed to have the same risk of death risk of Covid-19 patients. The results of this study indicate that the number of comorbids has a significant effect on the risk of death of Covid-19 patients, the more the number of comorbids suffered by the patient, the higher the risk of death of the patient. The accuracy of this method in classifying data is 84.68%.

Sex and Tobacco Use in Service Discontinuation in Rural Primary Care Settings

ABSTRACT Purpose Screening, brief intervention and referral to treatment in Alabama (SBIRT) integrates alcohol and substance use screenings in health care clinics in western Alabama. This study explores the relationships between sex and tobacco use with discontinuation of SBIRT services. Method Risk for alcohol and substance consumption was measured using the United States Alcohol Use Disorders Identification Test (US-AUDIT) and the Drug Abuse Screening Test (DAST-10). A convenience sample of 112 patients was utilized. A multivariable Bayesian binary regression model was estimated to determine risk factors for discontinuation. Results Overall, 64 (58.20%) individuals chose to discontinue services following their screening. Male tobacco users with positive screens for risky alcohol and/or drug use were more likely to discontinue services than non-tobacco users (aOR = 2.57, 95% CI = 1.11, 5.54). Discussion/ Conclusion SBIRT programs may need to address engagement and interventions specifically around tobacco use in relation to risky alcohol and substance use. Gender differences may reflect the need for individualized interventions.

The Modified-Half-Normal distribution: Properties and an efficient sampling scheme

We introduce a new family of probability distributions that we call the Modified-Half-Normal distributions. It is supported on the positive part of the real line with its probability density proportional to the function We explore a number of its properties including showing the fact that the normalizing constant and moments of the distribution can be represented in terms of the Fox-Wright function. We demonstrate its relevance by showing its connection to a number of Bayesian statistical methods appearing from multiple areas of research such as Bayesian Binary regression, Analysis of Directional data, and Bayesian graphical model. The availability of its efficient sampling scheme is important to the success of such Bayesian procedures. Therefore, a major focus of this article is the development of methods for generating random samples from the Modified-Half-Normal distribution. To ensure efficiency, we prove that the constructed accept reject algorithms are “uniformly efficient” with high acceptance probability irrespective of the choice of the parameter specifications. Finally, we provide basic inference procedure for its parameters when analyzing data assuming the Modified-Half-Normal probability model.

Modelagem Bayesiano de regressão binária para dados desbalanceados usando novas ligações

In this work, we presented, in a didactic way, the Bayesian binary regression modeling for unbalanced data using new links functions. Under the Bayesian approach and using information criteria, predictive evaluation measures and introducing the analysis of residuals, we show that the models that use power and reverse power link functions are better than traditional models in the presence of unbalanced data, considering two applications. Additionally, codes with the procedures presented using the Stan package are made available in &nbsp;order to facilitate the use of these models. The work also contains a simulation study that shows how the unbalance in the response variable affects the estimation of the parameters of a logistic regression with respect to the bias, mean square error and standard deviation of the estimates, regardless of the sample size. At the same time, considering two applications, we show how binary regression models with the power and reverse power links recently formulated in the literature can be used to adequately estimate the parameters in the type of unbalance considered.

Practice of Blood Donation and Associated Factors among Adults of Gondar City, Northwest Ethiopia: Bayesian Analysis Approach.

BackgroundBlood transfusion is an essential part of modern health care. In Ethiopia, it had been planned to collect 202,000 units of blood in 2016–2017, but the actual amount collected was 169,744. The Bayesian approach has an advantage in that estimation of model parameters is conducted based on posterior distribution. This study aimed to assess the practice of blood donation and associated factors among adults of Gondar city, northwest Ethiopia using Bayesian analysis.MethodsA community-based cross-sectional study was conducted among 554 Gondar adults from February to March 2019. Bayesian binary logistic regression was conducted to assess the relationship between independent and dependent variables using Stata 15 with a 95% CI for statistical significance.ResultsA total of 515 were involved with a 97% response rate, among which 342 (66.4%) were females. Eighty (15.53%) had donated blood at least once in their lives. Men were more likely to donate blood (AOR 1.79, 95% CI 1.11–3.12), while those aged 18–24 years were 57% (AOR 0.43, 95% CI 0.15–0.89) less likely to give blood. Individuals with good knowledge (AOR 2.56, 95% CI 1.32–4.53) and favorable attitudes (AOR 1.86, 95% CI 1.01–3.06) were more likely to donate blood.ConclusionThe majority of the participants did not donate blood. Male sex, age 18–24 years, good knowledge, and favorable attitudes were statistically significant factors in blood donation. Intervention with females and younger age-groups should shall be better administered. Health education is required to increase knowledge and create a favorable attitudes among the community.

Comparative analysis of Bayesian quantile regression models for pedestrian injury severity at signalized intersections

This study intended to (1) investigate the pedestrian injury severity involved in traffic crashes; and (2) address the heterogeneity issue at signalized intersections. To achieve the objectives, Bayesian binary and ordinal quantile regression models were proposed to address the pedestrian injury severity at signalized intersections. The suitability of the proposed method was illustrated with the Hong Kong dataset from 2008 to 2012 and 376 signalized intersections involving 2090 pedestrian-related crashes are selected. It’s found that age, injury location, pedestrian special circumstance, pedestrian contributory and the presence of Tram/LRT stops and right turning pocket are significant variables. The results indicated that both Bayesian binary and ordinal quantile regression models not only provide a more comprehensive and in-depth understanding of the relationship between pedestrian injury severity and the explanatory variables, but also highlight the heterogeneity issue for the data collected at different locations and different times without many assumptions. The goodness-of-fit of the proposed models outperforms existing mean models, while the Bayesian binary quantile model provides a better fit than the Bayesian quantile regression for ordinal model. The results can benefit the pedestrian facilities improvement/management and guide a much safer pedestrian environment.

Concomitant dysregulation of the estrogen receptor and BRAF/MEK signaling pathways is common in colorectal cancer and predicts a worse prognosis.

Recurrence is a major cause of colorectal cancer (CRC)-related death. As yet, the accurate identification of CRC patients at high risk of recurrence is still a major clinical challenge. Previously, we found that an estrogen receptor (ER) pathway gene signature may predict disease recurrence in CRC patients. The aim of this study is to evaluate the potential application of additional pathway-specific gene signatures in the prediction of CRC recurrence. The activities of 26 cancer-related pathways in CRC were semi-quantified using gene signature-based Bayesian binary regression analysis, and putative associations of the pathways with cancer recurrence risk were assessed using survival analysis. Among the 26 pathways tested, inactivation of the estrogen receptor (ER) pathway was found to be one of the most common events in CRC. Inactivation of this pathway was found to be frequently accompanied by over-activation of the BRAF/MEK pathway, and these two pathways were found to be associated with opposite effects on several clinicopathological CRC features, including microsatellite instability, subsite location, advanced stage and recurrence. Survival analysis of four independent CRC patient cohorts revealed that while the BRAF/MEK pathway was more strongly associated with recurrence than the ER pathway in mixed-stage CRCs, the ER pathway was a better predictor of recurrence than the BRAF/MEK pathway in stage II CRC. A combined use of these two pathways improved the prediction of CRC recurrence in both mixed stage CRC (n=1122; overall HR: 2.518, 95% CI: 1.570-4.038, p<0.001) and stage II CRC (n=535; overall HR: 1.976, 95% CI: 1.306-2.989, p=0.001). Combined activity of the ER and BRAF/MEK pathways may represent a novel biomarker for CRC prognosis and clinical management.

A Bayesian Binary Logistic Regression Approach in Identifying Factors Associated with Exclusive Breastfeeding Practices at Arba Minch Town, South Ethiopia

A Bayesian Binary Logistic Regression Approach in Identifying Factors Associated with Exclusive Breastfeeding Practices at Arba Minch Town, South Ethiopia

On The Bernoulli Mixture Model for Bidikmisi Scholarship Classification with Bayesian MCMC

This research has a purpose to develop Bernoulli Mixture model for Bidikmisi data modelling using Bayesian approach. Model development is done by considering the specificity in the data acceptance of Bidikmisi scholarship prototype in East Java Province. Bidikmisi acceptance status having a binary type (0 and 1) coupled with the main criteria factor of parent income and the number of dependents family produces a structure of Bernoulli mixture distribution with two components. The characteristics of each component can be identified through the Bernoulli Mixture modelling by involving the covariates of Bidikmisi scholarship recipients. The estimating parameter was performed using Bayesian Markov Chain Monte Carlo (MCMC) couple with the Gibbs Sampling algorithm. This model is applied to data registrants Bidikmisi districts/cities in the province of East Java as many as 44,489 students. This model shows the smallest value of Deviance Information Criteria (DIC) compared with Bayesian binary logistic regression.

Distinguishing real from fake ivory products by elemental analyses: A Bayesian hybrid classification method

As laws tighten to limit commercial ivory trading and protect threatened species like whales and elephants, increased sales of fake ivory products have become widespread. This study describes a method, handheld X-ray fluorescence (XRF) as a noninvasive technique for elemental analysis, to differentiate quickly between ivory (Asian and African elephant, mammoth) from non-ivory (bones, teeth, antler, horn, wood, synthetic resin, rock) materials. An equation consisting of 20 elements and light elements from a stepwise discriminant analysis was used to classify samples, followed by Bayesian binary regression to determine the probability of a sample being ‘ivory’, with complementary log log analysis to identify the best fit model for this purpose. This Bayesian hybrid classification model was 93% accurate with 92% precision in discriminating ivory from non-ivory materials. The method was then validated by scanning an additional ivory and non-ivory samples, correctly identifying bone as not ivory with >95% accuracy, except elephant bone, which was 72%. It was less accurate for wood and rock (25–85%); however, a preliminary screening to determine if samples are not Ca-dominant could eliminate inorganic materials. In conclusion, elemental analyses by XRF can be used to identify several forms of fake ivory samples, which could have forensic application.

A Bayesian supervised dual-dimensionality reduction model for simultaneous decoding of LFP and spike train signals.

Neuroscientists are increasingly collecting multimodal data during experiments and observational studies. Different data modalities-such as EEG, fMRI, LFP, and spike trains-offer different views of the complex systems contributing to neural phenomena. Here, we focus on joint modeling of LFP and spike train data, and present a novel Bayesian method for neural decoding to infer behavioral and experimental conditions. This model performs supervised dual-dimensionality reduction: it learns low-dimensional representations of two different sources of information that not only explain variation in the input data itself, but also predict extra-neuronal outcomes. Despite being one probabilistic unit, the model consists of multiple modules: exponential PCA and wavelet PCA are used for dimensionality reduction in the spike train and LFP modules, respectively; these modules simultaneously interface with a Bayesian binary regression module. We demonstrate how this model may be used for prediction, parametric inference, and identification of influential predictors. In prediction, the hierarchical model outperforms other models trained on LFP alone, spike train alone, and combined LFP and spike train data. We compare two methods for modeling the loading matrix and find them to perform similarly. Finally, model parameters and their posterior distributions yield scientific insights.

Bayesian binary quantile regression for the analysis of Bachelor-to-Master transition

ABSTRACTThe multi-cycle organization of modern university systems stimulates the interest in studying the progression to higher level degree courses during the academic career. In particular, after the achievement of the first level qualification (Bachelor degree), students have to decide whether to continue their university studies, by enrolling in a second level (Master) programme, or to conclude their training experience. In this work we propose a binary quantile regression (BQR) approach to analyse the Bachelor-to-Master transition phenomenon with the adoption of the Bayesian inferential perspective. In addition to the traditional predictors of academic outcomes, such as the personal characteristics and the field of study, different aspects of student's performance are considered. Moreover, the role of a new contextual variable, representing the type of university regulations experienced during the academic path, is investigated. The utility of the Bayesian BQR to characterize the non-continuation decision after the first cycle studies is illustrated with an application to administrative data of Bachelor graduates at the School of Economics of Sapienza University of Rome. The method favourably compares with more conventional model specifications concerning the conditional mean of the binary response.