Stable Conformal Prediction Sets

Abstract

When one observes a sequence of variables (x1, y1), ..., (xn, yn), conformal prediction is a methodology that allows to estimate a confidence set for yn+1 given xn+1 by merely assuming that the distribution of the data is exchangeable. While appealing, the computation of such set turns out to be infeasible in general, e.g., when the unknown variable yn+1 is continuous. In this paper, we combine conformal prediction techniques with algorithmic stability bounds to derive a prediction set computable with a single model fit. We perform some numerical experiments that illustrate the tightness of our estimation when the sample size is sufficiently large.