User:Lahrcr/sandbox

The Naïve 1 method led to sensible forecasts. It followed the in sample very well, which is expected. In the holdout out of sample, it also performed reasonably well because the residuals are not very large. In the out of sample forecast, the predictions seem reasonable given the recent course and past observations. One downfall of these predictions may be that it does not incorporate and seasonality into the predictions but rather stays flat at the last observed value.