In the page of Estimation , there are prediction interval of the normal distribution. In the similar way, there is the prediction interval of Regression Analysis .
For example, "If X = 10, Y = 25 is OK, Y = 30 is NG."
Gaussian Process Regression Model is used if we do not need to consider the formulation politely, and we need to get precise prediction, and we want to use the information of the density of the data.
Examples of R is in the page, Analysis of prediction interval by R.
NEXT 
Measurement Errors in Regression Analysis
