Find out how to use Conformalized Quantile Regression
We should know the way certain our mannequin is about its predictions to make well-informed selections. Therefore, returning solely some extent prediction isn’t sufficient. It doesn’t inform us something about whether or not we will belief our mannequin or not. If you wish to know why, take a look at my article beneath.
Within the article, I take advantage of a classification downside for instance. Nonetheless, many real-world issues are regression issues. For instance, we need to know the way sure a mannequin is when predicting tomorrow’s temperature.
Because the temperature is a steady variable, we want to know by which interval the true temperature will lie.
The broader the interval, the extra unsure the mannequin. Therefore, we should always belief it much less when making selections.
Two approaches come to thoughts. Both we use a set of fashions that predict the interval or we flip some extent prediction right into a…