Make predictions

1.

One of the best ways to evaluate a model is to use it to make a prediction. For each of the three models, generate a prediction of the run time for a problem size twice as big as the biggest data point, then run the experiment and see which model was most accurate.