2016/04/25 by Md Hasinur Rahaman Khan, Khan, Md Hasinur Rahaman, Anamika Bhadra +3 · 1 citation
Mathematics · #Advanced Causal Inference Techniques #FOS: Computer and information sciences #Methodology (stat.ME) #Statistical Methods and Bayesian Inference #Statistical Methods and Inference
paper · pdf · doi:10.48550/arxiv.1604.07311
openalex publication_date 2016/04/25 · openalex created_date 2022/10/03 · openalex updated_date 2026/07/28
The instability in the selection of models is a major concern with data sets\ncontaining a large number of covariates. We focus on stability selection which\nis used as a technique to improve variable selection performance for a range of\nselection methods, based on aggregating the results of applying a selection\nprocedure to sub-samples of the data where the observations are subject to\nright censoring. The accelerated failure time (AFT) models have proved useful\nin many contexts including the heavy censoring (as for example in cancer\nsurvival) and the high dimensionality (as for example in micro-array data). We\nimplement the stability selection approach using three variable selection\ntechniques--Lasso, ridge regression, and elastic net applied to censored data\nusing AFT models. We compare the performances of these regularized techniques\nwith and without stability selection approaches with simulation studies and a\nbreast cancer data analysis. The results suggest that stability selection gives\nalways stable scenario about the selection of variables and that as the\ndimension of data increases the performance of methods with stability selection\nalso improves compared to methods without stability selection irrespective of\nthe collinearity between the covariates.\n