In data regression, the standard error of estimate is defined by the square root of the summation of the difference of the predicted and actual values, divided by the number of data. Qualitatively, the sest measures the accuracy of the predicted values of your empirical model.