Showing papers by "Mei-Ling Ting Lee published in 2014"
TL;DR: This paper presents an alternative method to the existing regularization algorithm based on the estimates of the correlation matrices which minimize the mean squared error risk function and shows that it is more stable and provides more accurate results than the standard regularized canonical correlation method.
50 citations
TL;DR: This work adopted a two-part model to describe the overall survival experience for interval censored data with a cured proportion and constructed a BIC-type model selection method to recommend an appropriate specification of parametric and nonparametric components in the model.
Abstract: Varying-coefficient models have claimed an increasing portion of statistical research and are now applied to censored data analysis in medical studies. We incorporate such flexible semiparametric regression tools for interval censored data with a cured proportion. We adopted a two-part model to describe the overall survival experience for such complicated data. To fit the unknown functional components in the model, we take the local polynomial approach with bandwidth chosen by cross-validation. We establish consistency and asymptotic distribution of the estimation and propose to use bootstrap for inference. We constructed a BIC-type model selection method to recommend an appropriate specification of parametric and nonparametric components in the model. We conducted extensive simulations to assess the performance of our methods. An application on a decompression sickness data illustrates our methods.
23 citations
TL;DR: It is found that African Americans and more obese participants in MOST have a greater risk of developing severe knee OA.
Abstract: On the basis of longitudinal Kellgren-Lawrence (KL) grades of knee radiographs obtained from the Multicenter Osteoarthritis Study (MOST), we examine the association of obesity and race with severity of knee osteoarthritis (OA). We use the proportional odds model with mixed effects to conduct the analysis. Repeated KL grades were modeled as ordinal longitudinal measures, and a random effect term was included to adjust for the within-subject correlation among the KL grades over time. We found that African Americans and more obese participants in MOST have a greater risk of developing severe knee OA.
29 Sep 2014
TL;DR: Different types of lifetime data, methods for estimating hazard and survival functions, methodsfor risk assessment, as well as multivariate methods for survival data are reviewed.
Abstract: In this article we review several methods used in analyzing lifetime data. Lifetime data have found many applications in biomedical research, engineering, as well as in social sciences. Lifetime data can be defined as the times to the occurrence of a certain event. Statistical methods for analyzing lifetime data include the definition of survival time, censored data, prediction of probability of survival, prediction of instantaneous hazards, and evaluation of cure rates. We review different types of lifetime data, methods for estimating hazard and survival functions, methods for risk assessment, as well as multivariate methods for survival data. Lifetime models based on stochastic processes hitting a boundary, such as threshold regression for risk assessment, are also discussed.
Keywords:
hazard function;
survival function;
censored data;
failure-time models;
first-hitting-time models;
threshold regression