Abstract
It is observed that the predictivity is larger when predicting the change in absolute magnitude than when predicting the actual value in the weather forecasting, such as daily temperature and monthly rainful are easier to be predicted than the actual value. In this thesis, we investigate this observation by analyzing two specific stationary processes which are often used for modeling different physical phenomena, namely, the stationary AR(1) process and the stationary AR(2) process.
Post a Comment