Fused Lasso with the Adaptation of Parameter Ordering in Combining Multiple Studies with Repeated Measurements

Abstract

Summary Combining multiple studies is frequently undertaken in biomedical research to increase sample sizes for statistical power improvement. We consider the marginal model for the regression analysis of repeated measurements collected in several similar studies with potentially different variances and correlation structures. It is of great importance to examine whether there exist common parameters across study-specific marginal models so that simpler models, sensible interpretations and meaningful efficiency gain can be obtained. Combining multiple studies via the classical means of hypothesis testing involves a large number of simultaneous tests for all possible subsets of common regression parameters, in which it results in unduly large degrees of freedom and low statistical power. We develop a new method of fused lasso with the adaptation of parameter ordering (FLAPO) to scrutinize only adjacent-pair parameter differences, leading to a substantial reduction for the number of involved constraints. Our method enjoys the oracle properties as does the full fused lasso based on all pairwise parameter differences. We show that FLAPO gives estimators with smaller error bounds and better finite sample performance than the full fused lasso. We also establish a regularized inference procedure based on bias-corrected FLAPO. We illustrate our method through both simulation studies and an analysis of HIV surveillance data collected over five geographic regions in China, in which the presence or absence of common covariate effects are reflective to relative effectiveness of regional policies on HIV control and prevention.

EPS-LASSO: test for high-dimensional regression under extreme phenotype sampling of continuous traits

“…Our method is motivated by the general theory of hypothesis test for high-dimensional models, which answers the question by dealing with the score statistic in high-dimension. There is another de-biased technique that decomposes the estimate of regression coefficients into a bias term and a normally distributed term, which facilitates the derivation of Wald statistics (Javanmard and Montanari, 2014b;van de Geer et al, 2014;Wang et al, 2016). In our method, the decorrelated score function can be regarded as an approximately unbiased estimation function for b (Godambe and Kale, 1991).…”

A Distributed and Integrated Method of Moments for High-Dimensional Correlated Data Analysis

“…DIMM utilizes the full strength of GMM to combine information from multiple sources to achieve greater statistical power, an approach that has been shown to work well with longitudinal data; see for examples Wang et al (2012) and Wang et al (2016). DIMM has the potential to combine multimodal data, an important analytic task in biomedical data analysis for personalized medicine.…”

Adaptive Contrast Weighted Learning for Multi-Stage Multi-Treatment Decision-Making

“…The key to identifying the optimal treatment regime lies in the estimation of and . Wang, Wang, and Song () show that given root‐ n consistent estimators , the corresponding orders are also consistent. An intuitive approach is to posit a parametric regression model for to get the regression estimator , and then we can obtain directly from .…”