Adaptive thresholding of wavelet coefficients

Controlling the false discovery rate in behavior genetics research

“…In many applied problems it has been argued that the control of the FDR at some specified level is the more appropriate response to the multiplicity concern: in educational research [16], signal processing [1], Medical Research [12], in Psychology [11] and in Genetics [14]. The practical difference between the two approaches is not small or trivial, and the larger the problem the more dramatic the difference is.…”

“…Reject the hypothesis corresponding to P (1) if it is smaller than or equal to the threshold h (1) , and continue to reject the hypotheses as long as P (i ) is less than or equal to h (i ) . Stop when P (k) \ h (k) for the first time.…”

“…Therefore, the correct procedure to use in this example is the more general second procedure. 1 Two endpoints were added to the list of differences after using the FDR controlling procedure: the number of excursions, and dynamics of diversity. The difference discovered in the second endpoint is of special interest, as 'diversity' reflects the spatial and temporal spread of stops within a specified time window.…”

From Local Kernel to Nonlocal Multiple-Model Image Denoising

“…Here we wish to mention also the control of error rate thresholding developed by Abramovich and Benjamini [1], Benjamini and Liu [7], and Abramovich et al [2].…”

“…During the second stage, the estimateŷ from the Þrst stage is exploited also to improve the accuracy of the blockmatching. Details of the algorithm is documented in Dabov et al [19] and Matlab codes of the algorithm are available online 1 . In particular, in [19] we discuss and analyze the impact of the choice of the T 2D and T 1D transforms to the effectiveness of the algorithm.…”