mne_denoise.dss.denoisers.NonlinearDenoiser

class mne_denoise.dss.denoisers.NonlinearDenoiser

Base class for nonlinear/adaptive denoiser functions.

Nonlinear denoisers operate on source time series rather than sensor data. They are used in the iterative DSS algorithm where the denoising function is applied to the current source estimate at each iteration.

Examples include variance-based masking, kurtosis maximization, and other adaptive transformations.

References

Särelä & Valpola (2005). Denoising Source Separation. J. Mach. Learn. Res., 6, 233-272. Section 2.1 “One-Unit Algorithm for Source Separation”

__init__(*args, **kwargs)

Methods

__init__(*args, **kwargs)
denoise(source)	Apply nonlinear denoising to source time series.