Fused Lasso with the Adaptation of Parameter Ordering (Flapo) in Merging Multiple Studies with Repeated Measurements

• Activity Number: 240
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Learning and Data Mining
• Author(s): Fei Wang*+ and Lu Wang and Peter Song
• Companies: Ford Motor Credit and University of Michigan and University of Michigan
• Keywords: Data merging ; Error bounds ; Meta analysis

Abstract

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. Merging multiple studies using the classical means of hypothesis testing involves a large number of simultaneous tests, which probably suffer from low statistical power. We develop a new fused lasso method, using estimated parameter ordering, to scrutinize only adjacent-pair parameter differences, leading to a substantial reduction for the number of needed comparisons. Our method enjoys the oracle properties as does the full fused lasso that involves all pairwise parameter differences. We show that the proposed procedure has smaller error bounds and better finite sample performance than the full fused lasso.