Random threshold for linear model selection, revisited

In [11], a random thresholding method is introduced to select the significant, or non-null, mean terms among a collection of independent random variables, and applied to the problem of recovering the significant coefficients in nonordered model selection. We introduce a simple modification which removes the dependency of the proposed estimator on a window parameter while maintaining its asymptotic properties. A simulation study suggests that both procedures compare favorably to standard thresholding approaches, such as multiple testing or model-based clustering, in terms of the binary classification risk. An application of the method to the problem of activation detection on functional magnetic resonance imaging (fMRI) data is discussed.

Keywords

random threshold, non-ordered model selection, FDR, mixture modeling, binary risk, oracle risk, fMRI

Full Text (PDF format)

Published 15 May 2012