Three forecasting methods are considered and all three show comparable residual diagnostic results. Mean Absolute Percent Error (MAPE) is calculated In-sample (training dataset with 240 time periods), Out-of-sample (testing dataset with 60 time periods) and All data (training + testing dataset with a total of 300 time periods) and provided in the table below.

	In-Sample MAPE	Out-of-Sample MAPE	Training + Testing data
Forecasting Method 1	6.78%	8.32%	7.09%
Forecasting Method 2	8.24%	7.53%	8.10%
Forecasting Method 3	7.14%		7.36%

1. Given the forecasting accuracy of the three methods above, which forecasting method would you recommend be used?