how is YMSE calculated

[YF,YMSE] = forecast(AR_model,step,data)

Could anyone guide me on how is YMSE calculated, I couldn't find any code or algorithm for calculating YMSE?

Thank you!

Answer 1

[Shashank Gupta]

Hi,

forecast function explicitly does not return the error matrix, The function forecasts the output of an identified time series model in to some steps into the future. Although you can use compare function and find out the goodness of fit of your predicted model. Alternately you can use your own error matrix to check if the forecast model working as intended.

I hope this helps.

Cheers.

[jialun chen]

Thank you for your answer. I am referring to this website. Error variances of forecasted observations (ymse) was calculated by this forecast function. However, I don't understand how he calculated this.

[Shashank Gupta]

YMSE - Forecast error variances of future observations

For time-invariant models in which the length of each observation vector (n) is the same, this is a numPeriods-by-n matrix. For time-varying models in which the length of the observation vector changes, this is a numPeriods-by-1 cell array in which each element contains a time-varying n-element vector of forecast error variances associated with the corresponding period.

If you want to know more information for the function, You can access the code simply by

edit forecast