Can I calculate prediction intervals as a one off for the paper?
This can be done by using the final model for each metric, and bootstrapping the data, putting it through the model 100 times, and coming up with some sort of interval/measure of prediction interval
Can I calculate prediction intervals as a one off for the paper? This can be done by using the final model for each metric, and bootstrapping the data, putting it through the model 100 times, and coming up with some sort of interval/measure of prediction interval