- Browse by Author
Browsing by Author "Li, Xiaochen"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
Item Modern Monte Carlo Methods and Their Application in Semiparametric Regression(2021-05) Thomas, Samuel Joseph; Tu, Wanzhu; Boukai, Ben; Li, Xiaochen; Song, FengguangThe essence of Bayesian data analysis is to ascertain posterior distributions. Posteriors generally do not have closed-form expressions for direct computation in practical applications. Analysts, therefore, resort to Markov Chain Monte Carlo (MCMC) methods for the generation of sample observations that approximate the desired posterior distribution. Standard MCMC methods simulate sample values from the desired posterior distribution via random proposals. As a result, the mechanism used to generate the proposals inevitably determines the efficiency of the algorithm. One of the modern MCMC techniques designed to explore the high-dimensional space more efficiently is Hamiltonian Monte Carlo (HMC), based on the Hamiltonian differential equations. Inspired by classical mechanics, these equations incorporate a latent variable to generate MCMC proposals that are likely to be accepted. This dissertation discusses how such a powerful computational approach can be used for implementing statistical models. Along this line, I created a unified computational procedure for using HMC to fit various types of statistical models. The procedure that I proposed can be applied to a broad class of models, including linear models, generalized linear models, mixed-effects models, and various types of semiparametric regression models. To facilitate the fitting of a diverse set of models, I incorporated new parameterization and decomposition schemes to ensure the numerical performance of Bayesian model fitting without sacrificing the procedure’s general applicability. As a concrete application, I demonstrate how to use the proposed procedure to fit a multivariate generalized additive model (GAM), a nonstandard statistical model with a complex covariance structure and numerous parameters. Byproducts of the research include two software packages that all practical data analysts to use the proposed computational method to fit their own models. The research’s main methodological contribution is the unified computational approach that it presents for Bayesian model fitting that can be used for standard and nonstandard statistical models. Availability of such a procedure has greatly enhanced statistical modelers’ toolbox for implementing new and nonstandard statistical models.Item Natural History Of Implantable Cardioverter-Defibrillator Implanted At Or After The Age Of 70 Years In A Veteran Population A Single Center Study(2016-12) Ajam, Tarek; Kalra, Vikas; Shen, Changyu; Li, Xiaochen; Gautam, Sandeep; Kambur, Thomas; Barmeda, Mamta; Yancey, Kyle W.; Ajam, Samer; Garlie, Jason; Miller, John M.; Jain, Rahul; Medicine, School of MedicineBackground: The median age of patients in major Implantable Cardioverter-defibrillator (ICD)trials (MUSTT, MADIT-I, MADIT-II, and SCD-HeFT) was 63-67 years; with only 11% ≥70 years. There is little follow-up data on patients over 70 years of age who received an ICD for primary/secondary prevention of sudden cardiac death, particularly for veterans. Objective: The aim of this study was to study the natural history of ICD implantation for veterans over 70 years of age. Methods: We retrospectively reviewed single center ICD data in 216 patients with a mean age at implantation 76 ± 4 years. The ICD indication was primary prevention in 161 patients and secondary prevention in 55 patients. The ICD indication was unavailable in 4 patients. Results: Mean duration of follow up was 1686 ± 1244 days during which 114 (52%) patients died. Of these, 31% died without receiving any appropriate ICD therapy. Overall, 60/216 (28%) received appropriate therapy and 28/216 (13%) received inappropriate therapy. Patients who had ICD implantation for secondary prophylaxis had statistically more (p= 0.02) appropriate therapies compared to patients who had ICD implantation for primary prevention. Indication for implantation and hypertension predicted appropriate therapy, while age at the time of implantation and presence of atrial fibrillation predicted inappropriate therapies. Overall, 7.7% had device related complications. Conclusions: Although 28% septuagenarians in this study received appropriate ICD therapy, they had high rates of mortality, inappropriate therapy, and device complications. ICD implantation in the elderly merits individualized consideration, with higher benefit for secondary prevention.Item Subgroup Identification in Clinical Trials(2020-04) Li, Xiaochen; Gao, Sujuan; Shen, Changyu; Boukai, Ben; Zhang, Jianjun; Liu, HaoSubgroup analyses assess the heterogeneity of treatment effects in groups of patients defined by patients’ baseline characteristics. Identifying subgroup of patients with differential treatment effect is crucial for tailored therapeutics and personalized medicine. Model-based variable selection methods are well developed and widely applied to select significant treatment-by-covariate interactions for subgroup analyses. Machine learning and data-driven based methods for subgroup identification have also been developed. In this dissertation, I consider two different types of subgroup identification methods: one is nonparametric machine learning based and the other is model based. In the first part, the problem of subgroup identification was transferred to an optimization problem and a stochastic search technique was implemented to partition the whole population into disjoint subgroups with differential treatment effect. In the second approach, an integrative three-step model-based variable selection method was proposed for subgroup analyses in longitudinal data. Using this three steps variable selection framework, informative features and their interaction with the treatment indicator can be identified for subgroup analysis in longitudinal data. This method can be extended to longitudinal binary or categorical data. Simulation studies and real data examples were used to demonstrate the performance of the proposed methods.Item Type 2 Diabetes Genetic Risk Scores Are Associated With Increased Type 2 Diabetes Risk Among African Americans by Cardiometabolic Status(Sage, 2018-01-03) Layton, Jill; Li, Xiaochen; Shen, Changyu; de Groot, Mary; Lange, Leslie; Correa, Adolfo; Wessel, Jennifer; Epidemiology, School of Public HealthThe relationship between genetic risk variants associated with glucose homeostasis and type 2 diabetes risk has yet to be fully explored in African American populations. We pooled data from 4 prospective studies including 4622 African Americans to assess whether β-cell dysfunction (BCD) and/or insulin resistance (IR) genetic variants were associated with increased type 2 diabetes risk. The BCD genetic risk score (GRS) and combined BCD/IR GRS were significantly associated with increased type 2 diabetes risk. In cardiometabolic-stratified models, the BCD and IR GRS were associated with increased type 2 diabetes risk among 5 cardiometabolic strata: 3 clinically healthy strata and 2 clinically unhealthy strata. Genetic risk scores related to BCD and IR were associated with increased risk of type 2 diabetes in African Americans. Notably, the GRSs were significant predictors of type 2 diabetes among individuals in clinically normal ranges of cardiometabolic traits.