Conformal prediction

Conformal prediction is a way of making predictions that helps us say how sure we are about those predictions. It's like if you were trying to guess how many candies are in a jar, but instead of just guessing one number, you guess a range of numbers that you think might be right.

So, instead of saying "There are 20 candies in the jar" and being either right or wrong, you might say "There are probably between 15 and 25 candies in the jar." This way, you're not 100% sure, but you're more likely to be right than if you just guessed one number.

This can be really helpful in lots of different situations where we're trying to make predictions. For example, if we're trying to predict the weather, we might say "It's probably going to rain sometime in the next 24 hours, but we're not sure exactly when." Or if we're trying to predict who will win a basketball game, we might say "It could be either team, but we think one team is slightly more likely to win than the other."

By using conformal prediction, we can be more careful with our predictions and make sure that we're not making claims that are too bold or overconfident. We can still make predictions, but we just do it in a more cautious and thoughtful way.