Linear fixed weight combination prediction model and model optimum seeking method

It is educed that synchronously considering the accuracy and stability of prediction model is the primary reason that the linear fixed weight combination prediction model based on minimizing error sum of squares is widely applied to differently domanial prediction problems now by analyzing. On this base, the linear fixed weight combination prediction model based on prediction model validity is built, and example analysis indicates that new model has higher prediction precision. Aim at the problems that the optimal criterion selection of linear fixed weight combination prediction model have not followed gist and exist many subjective factors, the linear fixed weight combination prediction model optimum seeking method based on prediction error estimation is put forward, and whose validity is validated by example.