Search

We propose an adaptive design for variable selection in Bayesian modeling process. First randomly select some models to evaluate (e.g. by posterior model probability). Using these models, we predict the performance of all models in the candidate pool, based on which more models are selected and evaluated, in which models with good predicted performance or large prediction variances have high probabilities of being selected. Newly sampled models are used to update the performance predictions of candidate models. Repeat the process until informative models are not likely to be left unsampled in terms of the preset model selection criterion. When there are high-dimensional variables, we propose the use of highest-resolution-minimum-aberration-fractional-factorial design to select candidate-model sets to enable inferences on main effects and low-level interactions of variables. Simulations and a real data example have shown that the proposed adaptive design is efficient in finding informative models compared with other variable selection procedures.

The need for data integration is becoming ubiquitous and encompasses many disciplines due to the technological development in instrumentation. Combining the information from distinct data sources in modeling, so as to improve the prediction accuracy and have a holistic view of the problem is a challenge for statisticians. In this paper, we present a flexible statistical framework for integrating various types of data from distinct sources through model-based boosting (IMBoost) with two types of base models: regression trees and penalized splines. The performance of IMBoost is illustrated through two recent studies in environmental soil science, where multiple sensors were used to quantify several soil parameters. Empirical results are promising and show the proposed algorithms substantially improve the prediction performance through combining the strength from distinct data sources. We also proposed a surrogate model approach, which allows IMBoost to handle situations when partial samples are missing from distinct sources.

Mediation analysis allows the examination of effects of a third variable (mediator/confounder) in the causal pathway between an exposure and an outcome. The general multiple mediation analysis method (MMA), proposed by Yu et al., improves traditional methods (e.g., estimation of natural and controlled direct effects) to enable consideration of multiple mediators/confounders simultaneously and the use of linear and nonlinear predictive models for estimating mediation/confounding effects. Previous studies find that compared with non-Hispanic cancer survivors, Hispanic survivors are more likely to endure anxiety and depression after cancer diagnoses. In this paper, we applied MMA on MY-Health study to identify mediators/confounders and quantify the indirect effect of each identified mediator/confounder in explaining ethnic disparities in anxiety and depression among cancer survivors who enrolled in the study. We considered a number of socio-demographic variables, tumor characteristics, and treatment factors as potential mediators/confounders and found that most of the ethnic differences in anxiety or depression between Hispanic and non-Hispanic white cancer survivors were explained by younger diagnosis age, lower education level, lower proportions of employment, less likely of being born in the USA, less insurance, and less social support among Hispanic patients.

Background The aim of this study was to identify any associations between predictor variables, mainly clinician training and dental implant outcome, among the residents in different departments and to compare statistical analysis with the use of MATLAB R2017a™ to SAS version 9.4. Methods Dental records were reviewed from January 1, 2011, to December 31, 2015. Two thousand forty-eight dental implants were placed on 471 patients seen by residents from the departments of Periodontics, Prosthodontics, and Oral and Maxillofacial Surgery (OMFS) at Louisiana State University Health Sciences Center School of Dentistry. The following parameters were investigated by means of multilevel logistic regression analysis: demographics, implant parameters, department, and residents’ year of training. Results A total of 1449 implants were included in the study. Overall, within a 1–5-year time period, 1343 (92.6%) implants had survived and 106 (7.4%) implants failed. Discipline (p = 0.0004), residents’ year of training (p