Next: The basic LSL algorithm:
Up: Darche: Adaptive deconvolution
Previous: Introduction
I present here first the basic LSL algorithm as it is used for spectral
analysis: its formalism, the definitions of the forward and backward
residuals, and the recursion formulas. The proofs for the recursions implied
are long, and heavy: I prefer to refer to Lee et al. (1981) for elegant
geometric demonstrations. The recursion formulas also provide us with
a statistical tool, describing the time-statistical variations in the data,
which I will briefly explain.
Finally, I derive a simple version of the general LSL algorithm; Friedlander
(1982) had indeed derived such an algorithm, but in a ``normalized''
formalism, not easy to understand intuitively.
Next: The basic LSL algorithm:
Up: Darche: Adaptive deconvolution
Previous: Introduction
Stanford Exploration Project
1/13/1998