Nollakeskiarvolla
Nollakeskiarvolla, a Finnish term meaning with zero mean, is used in statistics, data analysis, and signal processing to describe a random variable, a stochastic process, or a data sequence whose expected value or arithmetic average is zero. In probabilistic terms, a random variable X is zero-mean if the expectation E[X] equals zero; a time series {X_t} is zero-mean if E[X_t] = 0 for all t. For deterministic data, a dataset has zero mean when the arithmetic average of its values is zero.
The concept is central to data centering, where the sample mean is subtracted from observations to produce
In signal processing, a zero-mean signal lacks a constant offset. A zero-mean white-noise process has an average
Applications of the zero-mean concept extend from model diagnostics (where residuals are assumed to be zero-mean)
See also: centering, mean subtraction, DC component, zero-mean processes.