Bonus1 [5 Points] A forecast model composed of an ARMA model and an ES model is used to predict the future demand of an automotive spare part. The variance of the forecast error of the ARMA model is 16, while the variance of the forecast error of the ES model is 25. The coefficient of correlation between the errors of these two models is -0.9. Calculate The optimal values of weights given to each of these two models and the value of the error variance of the combined model