It means that if, for instance, there is a long series of missing data at one or more stations in the validation period and in the calibration period we have another gap pattern, then the general mean-square error will be shifted to the value at one of stations.

The difference between the mean-square error estimated at the validation period, which equals to 2.

n] are unavailable is based on the use of linearly constrained minimum

mean-square error methods.

2] square and a lower regression

mean-square error, it would typically be considered to be the superior model based on goodness-of-fit criteria in the estimation sample.

The most common difference measure is the mean-square error (MSE).

To estimate entire image mean-square error (MSE) is used

This approximation is reasonable when the excess mean-square error (EMSE) variance is less than 10% of the minimum mean--square error (MSE) variance [1-2].

The excess mean-square error of the adaptive weights updating algorithm, caused by the fluctuation in the weight coefficients, can be defined as: