[KongchangAI]
Concept累积量特征

Higher-Order Cumulants

A classical statistical tool in signal processing that describes the degree to which a random process deviates from a Gaussian distribution through higher-order statistics. Higher-order cumulants are identically zero for Gaussian processes, which gives them an inherent ability to suppress Gaussian noise and makes them useful for non-Gaussian signal analysis and blind source separation.

Timeline (last 90 days)

Nov 5

Higher-order cumulants of Gaussian noise are theoretically zero, so cumulant features have a natural suppression capability against additive white Gaussian noise.

Unverified50%

All Facts (1)

Source Articles