[Python] How to let an AI model chooser whether to make a prediction or not for each data in a...

I have a dataset with a lot of data and I use sklearn to make a regressor model to get a number for each data (line) in the set.

It works fine but I want the model to make a prediction only if it is confident enough that the said prediction will be accurate. (Also being able to tweak what the model consider accurate enough will be neat.)

I thought about having the model predict a boolean and only consider the number if it is True, but I believe there must be a 'cleaner' way of doing it (with a dedicated method or something).

How can I do this ?

Continue reading...