Self-calibrating Probability Forecasting

Part of Advances in Neural Information Processing Systems 16 (NIPS 2003)

Bibtex Metadata Paper


Vladimir Vovk, Glenn Shafer, Ilia Nouretdinov


In the problem of probability forecasting the learner’s goal is to output, given a training set and a new object, a suitable probability measure on the possible values of the new object’s label. An on-line algorithm for probability forecasting is said to be well-calibrated if the probabilities it outputs agree with the observed frequencies. We give a natural non- asymptotic formalization of the notion of well-calibratedness, which we then study under the assumption of randomness (the object/label pairs are independent and identically distributed). It turns out that, although no probability forecasting algorithm is automatically well-calibrated in our sense, there exists a wide class of algorithms for “multiprobability forecasting” (such algorithms are allowed to output a set, ideally very narrow, of probability measures) which satisfy this property; we call the algorithms in this class “Venn probability machines”. Our experimental results demonstrate that a 1-Nearest Neighbor Venn probability machine performs reasonably well on a standard benchmark data set, and one of our theoretical results asserts that a simple Venn probability machine asymptotically approaches the true conditional probabilities regardless, and without knowledge, of the true probability measure generating the examples.