next up previous [pdf]

Next: Abma test Up: Fomel: Adaptive multiple suppression Previous: Nonstationary deconvolution

Adaptive multiple subtraction

In this section, I apply regularized nonstationary regression to the problem of adaptive subtraction and evaluate its performance using benchmark synthetic tests.

One possible variation of non-stationary regression applied to adaptive subtraction is an introduction of the signal prediction filter (Spitz, 1999). The signal prediction applies to the residual error $e(\mathbf{x})$ in equation 2, thus modifying the objective function (Guitton, 2005).