We propose a novel class of deep stochastic predictors for classifying metric data on graphs within the PAC-Bayes risk certification paradigm. Classifiers are realized as linearly parametrized deep assignment flows with random initial conditions. Building on the recent PAC-Bayes literature and data-dependent priors, this approach enables (i) to use risk bounds as training objectives for learning posterior distributions on the hypothesis space and (ii) to compute tight out-of-sample risk certificates of randomized classifiers more efficiently than related work. Comparison with empirical test set errors illustrates the performance and practicality of this self-certifying classification method.
@article{arxiv.2201.11162,
title = {Self-Certifying Classification by Linearized Deep Assignment},
author = {Bastian Boll and Alexander Zeilmann and Stefania Petra and Christoph Schnörr},
journal= {arXiv preprint arXiv:2201.11162},
year = {2022}
}