Binary Response Research Articles

Binary response of runoff generation and erosion processes to karst vegetation-soil-rock structure and its hydrodynamic mechanisms

Binary response of runoff generation and erosion processes to karst vegetation-soil-rock structure and its hydrodynamic mechanisms

Exploring Potential Adaptation Strategies for Smallholder Farmers to Climate Change: A Case Study at Abala Abaya, Wolaita Zone, South Ethiopia

Background: The frequency and intensity of extreme weather events are expected to increase as a result of climate change, which will contribute to changes in precipitation and temperature, and therefore will affect food security. Research issues related to increasing the sustainability of the agricultural sector, eliminating climate risks in agriculture are coming to the fore. This is especially relevant in arid regions. Objectives: This study aimed to assess the adaptation strategies of smallholder farmers to climate change in Abela Abaya, Wolaita Zone, Southern Ethiopia. Finding optimal practices adapted to random extreme events is fundamental for sustainable food production. In particular, the study attempted to understand the preferences of smallholder farmers in adaptation strategies to climate change in the study area; the main influencing factors determining the choice of adaptation strategies to climate change by smallholder farmers in the study area. Methods: To investigate this issue, qualitative and quantitative data were used within a descriptive research approach to assess potential adaptation strategies of smallholder farmers to climate change. Data on demographic, socio-economic, institutional, physical and psychological factors for adaptation strategies of smallholder farmers were collected through specially designed and pre-tested questionnaires, interviews and focus group discussions (FGDs). Rural counties for the study were selected using simple random sampling due to the same agroecology. The multivariate probit (MVP) model was used as it is a type of correlated binary response regression model that allows for the simultaneous identification and assessment of the impact of a set of independent factors on each of the possible approaches. Five adaptation methods are identified as dependent variables for the Multi-Variate Probit. These include Soil and Water Conservation (SWC) practices, use of drought-tolerant varieties, adjustment of planting dates, use of agroforestry, and implementation of water harvesting. The following were selected as independent variables for the current study. Results: Most farmers (96.7%) acknowledge the ongoing climate change and are concerned that their agriculture will not suffer under the new conditions. This is justified by the fact that farmers confirm their observations regarding the increase in air temperature, the increase in the frequency of plant diseases and the decrease in precipitation. At the same time, only 36.2% of the surveyed farmers reported that society is aware of the possible risks associated with global warming and is trying to adapt agricultural activities to new climatic conditions. In particular, the current study revealed for the first time that smallholder farmers in Abela Abaya primarily prefer approaches such as soil and water conservation (71.6%), adjustment of planting dates (59.4%), and agroforestry (44.5%). Conclusion: The study identified the most influential factors on the adoption of adapted agricultural practices in a specific region of Abala Abaya (Wolaita Zone, Southern Ethiopia), namely, household heads’ access to education, frequency of extension visits, access to climate information and land slope. Thus, decision makers can design and adopt appropriate programs based on the current results to preserve smallholder farming and maintain food security at the national level.

Flexible Bayesian modeling for longitudinal binary and ordinal responses

Longitudinal studies with binary or ordinal responses are widely encountered in various disciplines, where the primary focus is on the temporal evolution of the probability of each response category. Traditional approaches build from the generalized mixed effects modeling framework. Even amplified with nonparametric priors placed on the fixed or random effects, such models are restrictive due to the implied assumptions on the marginal expectation and covariance structure of the responses. We tackle the problem from a functional data analysis perspective, treating the observations for each subject as realizations from subject-specific stochastic processes at the measured times. We develop the methodology focusing initially on binary responses, for which we assume the stochastic processes have Binomial marginal distributions. Leveraging the logits representation, we model the discrete space processes through continuous space processes. We utilize a hierarchical framework to model the mean and covariance kernel of the continuous space processes nonparametrically and simultaneously through a Gaussian process prior and an Inverse-Wishart process prior, respectively. The prior structure results in flexible inference for the evolution and correlation of binary responses, while allowing for borrowing of strength across all subjects. The modeling approach can be naturally extended to ordinal responses. Here, the continuation-ratio logits factorization of the multinomial distribution is key for efficient modeling and inference, including a practical way of dealing with unbalanced longitudinal data. The methodology is illustrated with synthetic data examples and an analysis of college students’ mental health status data.

Prediction of hepatocellular carcinoma response to radiation segmentectomy using an MRI-based machine learning approach.

To evaluate the value of pre-treatment MRI-based radiomics in patients with hepatocellular carcinoma (HCC) for the prediction of response to Yttrium 90 radiation segmentectomy. This retrospective study included 154 patients (38 female; mean age 66.8years) who underwent contrast-enhanced MRI prior to radiation segmentectomy. Radiomics features were manually extracted on volumes of interest on post-contrast T1-weighted images at the portal venous phase (PVP). Tumor-based response assessment was evaluated 6months post-treatment using mRECIST. A logistic regression model was used to predict binary response outcome [complete response at 6months with no-re-treatment (response group) against the rest (non-response group, including partial response, progressive disease, stable disease and complete response after re-treatment within 6months after radiation segmentectomy) using baseline clinical parameters and radiomics features. We accessed the value of different sets of predictors using cross-validation technique. AUCs were compared using DeLong tests. A total 168 HCCs (mean size 2.9 ± 1.7cm) were analyzed in 154 patients. The response group consisted of 113 HCCs and the non-response group of 55 HCCs. Baseline clinical parameters (AUC 0.531; sensitivity, 0.781; specificity, 0.279; positive predictive value (PPV), 0.345; negative predictive value (NPV), 0.724) and AFP (AUC 0.632; sensitivity, 0.833; specificity, 0.466; PPV, 0.432; NPV, 0.851) showed poor performance for response prediction. The model using a combination of radiomics features and clinical parameters/AFP showed the best performance (AUC 0.736; sensitivity, 0.706; specificity, 0.662; PPV 0.504; NPV, 0.822), significantly better than the clinical model (p < 0.001) or AFP alone (p < 0.001). The combination of radiomics features from pre-treatment MRI with clinical parameters and AFP showed fair performance for predicting HCC response to radiation segmentectomy, better than that of AFP. These results need further validation.

Psychometric Assessment of the Temporal Bisection Task with Discrete and Continuous Response Formats

Abstract The temporal bisection task has long been used to study time perception as well as measure individual differences in time perception ability. The task involves training participants on short and long reference durations before presenting intermediate durations and asking participants to classify them as ‘short’ or ‘long’. However, there is little information about how well the bisection task measures individual differences in timing ability. To bridge this gap, we assessed the psychometric properties of measures obtained from a classic temporal bisection task: Weber ratio and percent correct. Because measures with binary responses tend to require many trials to reach adequate reliability, we also assessed the psychometric properties of a modified bisection task which used a continuous response format. In this task, participants represented intermediate durations on a visual analogue scale. Estimation error was used as the outcome measure. Participants (n = 46) completed the classic and modified bisection tasks twice across two sessions approximately one week apart. The modified bisection task had excellent internal consistency and test–retest reliability, while the classic task had fair to good internal consistency and good test–retest reliability. Overall, estimation error had the highest reliability, followed by percent correct, and then Weber ratio. In terms of validity, there was excellent convergent validity between the classic and modified bisection tasks. As an exploratory analysis, we assessed how the number of trials affected the reliability of each outcome measure across the two tasks. Based on this, we make recommendations on how to optimize reliability for both tasks in future research.

Model Averaging for Estimating Treatment Effects With Binary Responses

ABSTRACTIn this article, we present a novel approach for estimating the conditional average treatment effect in models with binary responses. Our proposed method involves model averaging, and we establish a weight choice criterion based on jackknife model averaging. We analyze the theoretical properties of this approach, including its asymptotic optimality in achieving the lowest possible squared error and the convergence rate of the weights assigned to correctly specified models. Additionally, we introduce a new matching method that combines partition and nearest neighbor pairing, leveraging the strengths of both techniques. To evaluate the performance of our method, we conduct comparisons with existing approaches via a Monte Carlo study and a real data analysis. Overall, our results demonstrate the effectiveness and practicality of our proposed approach for estimating the conditional average treatment effect in binary response models.

Risk Factors for Musculoskeletal Disorders in Korean Farmers: Survey on Occupational Diseases in 2020 and 2022.

This study investigated factors influencing the prevalence of musculoskeletal disorders (MSDs) resulting from agricultural work, utilizing the 2020 and 2022 occupational disease survey data collected by the Rural Development Administration. The combined data from these years indicated a 6.02% prevalence of MSDs, reflecting a significant class imbalance in the binary response variables. This imbalance could lead to classifiers overlooking rare events, potentially inflating accuracy assessments. We evaluated five distinct models to compare their performance using both original and synthetic data and assessing the models' performance based on synthetic data generation. In the multivariate logistic model, we focused on the main effects of the covariates as there were no statistically significant second-order interactions. Focusing on the random over-sampling examples (ROSE) method, gender, age, and pesticide use were particularly impactful. The odds of experiencing MSDs were 1.29 times higher for females than males. The odds increased with age: 2.66 times higher for those aged 50-59, 4.60 times higher for those aged 60-69, and 7.16 times higher for those aged 70 or older, compared to those under 50. Pesticide use was associated with 1.26 times higher odds of developing MSDs. Among body part usage variables, all except wrists and knees were significant. Farmers who frequently used their necks, arms, and waist showed 1.27, 1.11, and 1.23 times higher odds of developing MSDs, respectively. The accuracy of the raw method was high, but the ROSE method outperformed it for precision and F1 score, and both methods showed similar AUC.

A Bayesian hybrid method for the analysis of generalized linear models with missing-not-at-random covariates

Missing data handling is one of the main problems in modelling, particularly if the missingness is of type missing-not-at-random (MNAR) where missingness occurs due to the actual value of the observation. The focus of the current article is generalized linear modelling of fully observed binary response variables depending on at least one MNAR covariate. For the traditional analysis of such models, an individual model for the probability of missingness is assumed and incorporated in the model framework. However, this probability model is untestable, as the missingness of MNAR data depend on their actual values that would have been observed otherwise. In this article, we consider creating a model space that consist of all possible and plausible models for probability of missingness and develop a hybrid method in which a reversible jump Markov chain Monte Carlo (RJMCMC) algorithm is combined with Bayesian Model Averaging (BMA). RJMCMC is adopted to obtain posterior estimates of model parameters as well as probability of each model in the model space. BMA is used to synthesize coefficient estimates from all models in the model space while accounting for model uncertainties. Through a validation study with a simulated data set and a real data application, the performance of the proposed methodology is found to be satisfactory in accuracy and efficiency of estimates.

Environmental and additive genetic factors that affect artificial insemination success in Latxa dairy sheep breed

In the Latxa breed, flocks within the breeding program make use of high genetic value rams to inseminate their ewes once per year. Despite the relevance of the insemination results in the genetic progress and on farms’ productivity, external and genetic factors that affect artificial insemination (AI) success have not been up to now explored in this sheep population. For that, 135,351 edited AI records from 63,480 inseminated Latxa Cara Negra from Euskadi ewes, using 853 service rams, collected between 2000 and 2021, were used. The outcome of an AI event was treated as a binary response of either success or failure in becoming pregnant. To identify the environmental factors influencing the AI result, a multiple logistic regression was first calculated on a selection of variables related to the ewes, to the ram, and to other non-sex-specific aspects. With relevant variables detected, a threshold model, including the pedigree information of the inseminated ewes and the service rams, was used to estimate the genetic components of the trait in both sexes. Findings show that the AI success is higher in ewes who had the previous parturition from an AI event, in those with a lambing-AI interval longer than 210d, with 3 years and 2 lambings, with high prolificacy and in those having their first parturition at 1 year of age rather than at 2 or 3 years of age. In counterpart, the higher the milk produced in the nearest record before AI date, the poorer the AI results. Furthermore, relevant variability was linked to the year of insemination, the herd and the technician in charge of the insemination procedure. Regarding genetic parameters, heritability values, on the observable scale, and repeatability were 0.057 ± 0.004 and 0.204±0.007 in females and 0.009 ± 0.037 and 0.032 ± 0.002 in males, respectively. These results evidence that the AI success is under moderately low additive genetic control, with the surrounding environmental variables being the strongest controlling factor. An effort to enhance some farm management practices should be encouraged to efficiently improve reproductive results. Although genetic selection on AI success is viable, the genetic progress may be scarce. An option to improve reproduction by means of selective breeding might be the development of a multi-trait index.

Implication of Exclusive Breastfeeding in Early Childhood Dental Disorders: Large Cohort Evidence, US National Survey of Children Health.

Purpose/Objective: Frequent consumption of fermented carbohydrates and bottle feeding contribute to the development of early childhood caries (ECC). There are no substantial findings on the effects of breastfeeding patterns on oral health conditions in children in the United States. We aimed to assess the nexus between exclusive breastfeeding during the first 6 months and oral health disorders, namely toothache, decayed teeth, or unfilled cavities in early childhood. Materials and Methods: Design: Cross-sectional epidemiologic design of nationally representative data collected via telephone surveys in the United States was utilized. Setting: National Survey of Children's Health, 2012 Participants: Population-based random sample of parents/guardians of 24,655 children aged 6 months to 5 years. Main Outcome and Measure: The primary outcome was the presence of oral health disorders (OHD) in children, defined as the parent-reported occurrence of one or more of the following conditions in the past 12 months: toothache, decayed teeth, or an unfilled cavity. This was assessed through a single composite question in the survey, asking parents/guardians if their child had experienced any of these specific oral health issues within the previous year with a binary (yes/no) response option. Results: Among 24,655 children aged 6 months to 5 years, 2392 (9.7%) had experienced an oral health disorder (OHD) in the previous 12 months. In unadjusted analysis, children exclusively breastfed for 6 months were 28% less likely to present with OHD compared to never-breastfed children (OR = 0.72, 95% CI, 0.52-0.98). However, after adjusting for key factors, including maternal health and race/ethnicity, this association was no longer statistically significant (aOR = 1.11, 95% CI 0.79-1.57). Significant predictors of OHD included maternal health (aOR for fair/poor vs. excellent/very good maternal health = 1.79, 95% CI 1.08-2.69) and race/ethnicity, with a higher prevalence among non-Hispanic Black (12.9%) and multi-racial children (12.6%) compared to non-Hispanic White children (7.7%). Conclusions and Relevance: While exclusive breastfeeding for the first six months of life was not found to be a significant predictor of pediatric oral health outcomes after adjustment, maternal health and race/ethnicity were significant determinants of oral health disorders. These findings underscore the importance of comprehensive healthcare approaches that consider maternal well-being and socio-demographic factors. Future research should explore interventions targeting these factors to improve pediatric oral health outcomes.

Simple or complex statistical models: Non-traditional regression models with intuitive interpretations

The evolution of statistical modelling has historically been constrained by the practical limitations of computation; early statistical modelling favoured models which could feasibly be estimated. As increased mathematical complexity often implies more intricate computation, statistical models have grown both mathematically and computationally more complex. However, paradoxically, sometimes conceptually simpler models present more computational challenges than complex ones, and these have historically been neglected. In the case of binary responses, logistic regression models are the gold standard; covariates are modelled additively on the log-odds scale. In the case of time-to-event responses, the Cox proportional hazards regression model has additivity on the log-hazard rate scale. Both of these methods are computationally convenient, and yet the scales on which covariates are modelled are far from intuitive. We demonstrate the use of alternative models which are computationally more complex, yet feasible, but in which modelling is on a more interpretable scale. In the case of binary responses, a more intuitive alternative is log-binomial regression, in which modelling is on the log-relative risk scale. In the case of time-to-event responses, distributional regression enables modelling on the time scale. While logistic regression and proportional hazards regression qualitatively both deliver the same conclusions as the alternative models, log-binomial and distributional regression provide more interpretable coefficients, which are readily estimated.

Mosna reveals different types of cellular interactions predictive of response to immunotherapies and survival in cancer.

Spatially resolved omics enable the discovery of tissue organization of biological or clinical importance. Despite the existence of several methods, performing a rational analysis including multiple algorithms while integrating different conditions such as clinical data is still not trivial. To make such investigations more accessible, we developed mosna, a Python package to analyze spatial omics data with respect to clinical or biological data and to gain insight on cell interaction patterns or tissue architecture of biological relevance. mosna is compatible with all spatial omics methods, it leverages tysserand to build accurate spatial networks, and is compatible with Squidpy. It proposes an analysis pipeline, in which increasingly complex features computed at each step can be explored in integration with clinical data, either with easy-to-use descriptive statistics and data visualization, or by seamlessly training machine learning models and identifying variables with the most predictive power. mosna can take as input any dataset produced by spatial omics methods, including sub-cellular resolved transcriptomics (MERFISH, seqFISH) and proteomics (CODEX, MIBI-TOF, low-plex immuno-fluorescence), as well as spot-based spatial transcriptomics (10x Visium). Integration with experimental metadata or clinical data is adapted to binary conditions, such as biological treatments or response status of patients, and to survival data. We demonstrate the proposed analysis pipeline on two spatially resolved proteomic datasets containing either binary response to immunotherapy or survival data. mosna identifies features describing cellular composition and spatial distribution that can provide biological insight regarding factors that affect response to immunotherapies or survival.

Nonparametric Test for Logistic Regression with Application to Italian Enterprises’ Propensity for Innovation

In this work, a nonparametric method is proposed to jointly test the significance of the regression coefficient estimates in a logistic regression model and identify which explanatory variables are effective in predicting the binary response. The motivating example is related to the factors affecting the propensity of Italian Small Medium Enterprises (SMEs) to innovate. The explanatory variables of the model represent firms’ characteristics, such as size and age, and the possible effect of the sector of economic activity is taken into account by including a set of binary variables as control factors. The dependent variable indicates whether a company, in the period under study, introduced at least one product or process innovation. Therefore, it is also dichotomous, and the logistic regression model is appropriate for representing the relationship between explanatory variables and dependent variables. Specifically, the logit transformation of the firm’s propensity to innovate, i.e., the probability that a company randomly chosen from the population of Italian SMEs has introduced an innovation or, equivalently, the proportion of innovative companies among the Italian SMEs, is expressed as a linear function of the predictors (explanatory and control variables). The proposed test is based on the permutation approach and satisfies important statistical properties, proved in a simulation study. The test is more flexible and robust than the classic parametric approach, and is preferable to typical stepwise regression procedures for the selection of a parsimonious and effective model.

Ownership structure, board composition and CEO dominance: what does really matters to corporate deleveraging policies?

Purpose The role played by corporate governance mechanisms on corporate deleveraging policies has not been clarified. Empirical evidence is confined to developed economies, even with conflicting and inconclusive results. This paper aims to examine the role of corporate governance mechanisms, such as ownership structure, board composition and CEO dominance, in explaining corporate deleveraging policies. Design/methodology/approach Using a sample of listed Pakistani firms between 2010 and 2022, this study resorts to binary response models to examine the effects of governance mechanisms on firms’ decision to go debt-free. Findings A greater ownership concentration, institutional ownership and family ownership increase the propensity for zero leverage. Board gender diversity decreases the propensity for deleveraging policies, which seems to indicate that the presence of females reinforces the monitoring function of the board. Finally, lower managerial ownership or CEO dominance decreases the propensity toward zero leverage (interest convergence hypothesis), but higher managerial ownership or CEO dominance increases the propensity toward zero leverage (managerial entrenchment hypothesis). Practical implications Risk-averse managers who prefer to control a firm using little or no debt will find it easier to implement these financing policies in firms with greater ownership concentration and where institutional holders have a substantial stake. For shareholders, this study suggests that investing in firms with females on board reduces the risk of corporate deleveraging policies being adopted for entrenched reasons. Social implications The presence of females on board seems to decrease the propensity of managers to adopt opportunistic actions and may also contribute to enhancing human welfare and society in developing countries. Originality/value To the best of the authors’ knowledge, this is the first study considering the effect of board diversity on zero leverage. Another singularity is that this study exhibits a nonlinear relationship between managerial ownership and corporate deleveraging policy.

Gender-related measurement invariance on the Self-Reporting Questionnaire (SRQ-20) for global mental distress with older adults in Puerto Rico

BackgroundCommon mental disorders (CMD) vary by age, gender, and culture. This study: (1) examined the factor structure of the 20-item Self Reporting Questionnaire (SRQ-20) and (2) explored gender-related measurement invariance in the SRQ’s performance with older adults in Puerto Rico, a U.S. island territory and associate member of the UN Regional Commissions.MethodsWe merged data from two cross-sectional studies on mental health status and needs of older adults in Puerto Rico (N = 367). The first study was conducted in 2019, two years after Hurricane María devastated the island (N = 154); the second study, in 2021, assessed knowledge, attitudes and practices (KAP) concerning COVID-19 (N = 213). We used chi-square and t-tests to examine gender differences in each SRQ item and assessed internal consistency reliability with Cronbach’s alpha and McDonald’s omega (values > 0.70). We ran two CFA models, then multigroup CFA to test for gender-related measurement invariance. We used weighted least square mean and variance adjusted (WLSMV) estimation to account for the binary response options in the SRQ-20 and Mplus version 8.4 for analyses. There were no missing data for any SRQ-20 items.ResultsThe SRQ-20 had strong internal consistency reliability (α = 0.89; omega = 0.89). Female scores were higher than males scores (t = -2.159, p = .031). Both unidimensional and two-factor models fit the data well. We selected the more parsimonious unidimensional model, which is most widely used in practice. Standardized factor loadings were 0.548 to 0.823 and all were statistically significant (p < .001). We tested gender invariance with the one-factor model. Our findings did not support invariance.ConclusionWe favored the unidimensional model. First, the SRQ-20 was designed to assess global distress. Also, physical symptoms have both somatic and psychological components, so their co-occurrence makes a single-factor model more meaningful. Finally, since older adults experience more physical health problems, instruments that emphasize both types of distress may provide a more accurate measure than those that exclude somatic symptoms. Using the unidimensional model, the SRQ-20 was not invariant, meaning that it performed differently for male and female participants. Future studies of common mental disorders with older adults in Puerto Rico should consider using the SRQ-20 for research and practice and should determine appropriate threshold scores for men and women.

The beauty of tipping: customer decision-making during COVID-19 in African beauty salons

ABSTRACT Research has been limited to investigating tipping in other hospitality industries, while ignoring beauty salons. The objective of this study is to present tipping behaviours in the context of beauty salons in two ways: first, by examining tip predictors; and second, by looking at the amount customers are willing and able to offer as tips. We employed a binary response strategy and a limited dependent-variable approach to test several factors with the use of STATA 17. Results show that tipping depends on service quality, individual social approval, future expectations, religious beliefs, the educational level of the consumer and a customer’s income level. Likewise, the loss of work or sources of income caused by COVID-19 reduces the probability of tipping and, because of health measures and social distancing during COVID-19, people’s tipping decisions and tip amounts increase. Thus, beauty salon practitioners (customers, stylists and managers) should keep their services up-to-date to attract more tips, and customers should not feel embarrassed when they do not have the ability to tip.

Bayesian multiple change-points detection in autocorrelated binary process with application to COVID-19 infection pattern

Maintaining consistent model parameters in long-run experiments is challenging. We study the multiple change-points detection problem from this viewpoint for an autocorrelated sample path of a binary response variable. The probabilities of success and correlation structure are linked to explanatory or control variables using a generalized linear model. A Bayesian paradigm is introduced here to estimate the unknown number and locations of change-points and an adaptive RJMCMC algorithm is presented to generate pseudo-random samples from the posterior distributions. This approach allows researchers to incorporate the experts' ideas into the analysis of such data. The features of the proposed method are examined via some simulation studies. We also apply the presented method to COVID-19 data from four countries to detect the change-points in the infection procedures in the presence of some explanatory variables.

Unmconf : an R package for Bayesian regression with unmeasured confounders

The inability to correctly account for unmeasured confounding can lead to bias in parameter estimates, invalid uncertainty assessments, and erroneous conclusions. Sensitivity analysis is an approach to investigate the impact of unmeasured confounding in observational studies. However, the adoption of this approach has been slow given the lack of accessible software. An extensive review of available R packages to account for unmeasured confounding list deterministic sensitivity analysis methods, but no R packages were listed for probabilistic sensitivity analysis. The R package unmconf implements the first available package for probabilistic sensitivity analysis through a Bayesian unmeasured confounding model. The package allows for normal, binary, Poisson, or gamma responses, accounting for one or two unmeasured confounders from the normal or binomial distribution. The goal of unmconf is to implement a user friendly package that performs Bayesian modeling in the presence of unmeasured confounders, with simple commands on the front end while performing more intensive computation on the back end. We investigate the applicability of this package through novel simulation studies. The results indicate that credible intervals will have near nominal coverage probability and smaller bias when modeling the unmeasured confounder(s) for varying levels of internal/external validation data across various combinations of response-unmeasured confounder distributional families.

Consistent and Scalable Bayesian Joint Variable and Graph Selection for Disease Diagnosis Leveraging Functional Brain Network

We consider the joint inference of regression coefficients and the inverse covariance matrix for covariates in high-dimensional probit regression, where the predictors are both relevant to the binary response and functionally related to one another. A hierarchical model with spike and slab priors over regression coefficients and the elements in the inverse covariance matrix is employed to simultaneously perform variable and graph selection. We establish joint selection consistency for both the variable and the underlying graph when the dimension of predictors is allowed to grow much larger than the sample size, which is the first theoretical result in the Bayesian literature. A scalable Gibbs sampler is derived that performs better in high-dimensional simulation studies compared with other state-of-art methods. We illustrate the practical impact and utilities of the proposed method via a functional MRI dataset, where both the regions of interest with altered functional activities and the underlying functional brain network are inferred and integrated together for stratifying disease risk.

Contextual Dynamic Pricing with Strategic Buyers

Personalized pricing, which involves tailoring prices based on individual characteristics, is commonly used by firms to implement a consumer-specific pricing policy. In this process, buyers can also strategically manipulate their feature data to obtain a lower price, incurring certain manipulation costs. Such strategic behavior can hinder firms from maximizing their profits. In this article, we study the contextual dynamic pricing problem with strategic buyers. The seller does not observe the buyer’s true feature, but a manipulated feature according to buyers’ strategic behavior. In addition, the seller does not observe the buyers’ valuation of the product, but only a binary response indicating whether a sale happens or not. Recognizing these challenges, we propose a strategic dynamic pricing policy that incorporates the buyers’ strategic behavior into the online learning to maximize the seller’s cumulative revenue. We first prove that existing nonstrategic pricing policies that neglect the buyers’ strategic behavior result in a linear Ω ( T ) regret with T the total time horizon, indicating that these policies are not better than a random pricing policy. We then establish an O ( T ) regret upper bound of our proposed policy and an Ω ( T ) regret lower bound for any pricing policy within our problem setting. This underscores the rate optimality of our policy. Importantly, our policy is not a mere amalgamation of existing dynamic pricing policies and strategic behavior handling algorithms. Our policy can also accommodate the scenario when the marginal cost of manipulation is unknown in advance. To account for it, we simultaneously estimate the valuation parameter and the cost parameter in the online pricing policy, which is shown to also achieve an O ( T ) regret bound. Extensive experiments support our theoretical developments and demonstrate the superior performance of our policy compared to other pricing policies that are unaware of the strategic behaviors. Supplementary materials for this article are available online, including a standardized description of the materials available for reproducing the work.