A0757
Title: On the validity of conformal prediction for network data under non-uniform sampling
Authors: Robert Lunde - Washington University in St Louis (United States) [presenting]
Abstract: The properties of conformal prediction for network data are studied under various sampling mechanisms that commonly arise in practice but often result in a non-representative sample of nodes. These sampling mechanisms are interpreted as selection rules applied to a superpopulation, and the validity of conformal prediction conditional is studied based on an appropriate selection event. It is shown that the sampled subarray is exchangeable conditional on the selection event if the selection rule satisfies a permutation invariance property and a joint exchangeability condition holds for the superpopulation. The result implies the finite-sample validity of conformal prediction for certain selection events related to ego networks and snowball sampling. It is also shown that when data are sampled via a random walk on a graph, a variant of weighted conformal prediction yields asymptotically valid prediction sets for an independently selected node from the population.
