Four different time series models using the same data were used to make a prediction for next months sales. The results of different methods of computing forecast errors are presented in the following table. Bias MAD MSE MAPE Modell 3.9? 114.95 14.23 Model 2 3.12 6.34 109.32 11.29 Model 3 4131 10.23 113.49 13.94 Mode|4 3.06 3.62 112.31 13.9?r Using MAPE as the decision criteria the best model to select is l 59'3\"] Using MAD as the decision criteria the best model to select is l 59'3\"] Use the data below. January 20 February 21 March 15 April 14 May 13 June 16 July 17 August 18 September 20 October 20 November 21 December 23 A three month weighted moving average will predict sales of [ Select ] * for the month of January. With a MAD of [ Select ] * . Use the weights 1 for the observation furthest away in time, 2 for the middle observations and 3 for the closest observation in time