Towards a Unified Information-Theoretic Framework for Generalization

In this work, we investigate the expressiveness of the "conditional mutual information" (CMI) framework of Steinke and Zakynthinou [1] and the prospect of using it to provide a unified framework for proving generalization bounds in the realizable setting. We first demonstrate that one can use this framework to express non-trivial (but sub-optimal) bounds for any learning algorithm that outputs hypotheses from a class of bounded VC dimension. We then explore two directions of strengthening this bound: (i) Can the CMI framework express optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal optimal bounds for VC classes? (ii) Can the CMI framework be used to analyze algorithms whose output hypothesis space is unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted unrestricted (i.e. has an unbounded VC dimension)? With respect to Item (i) we prove that the CMI framework yields the optimal bound on the expected risk of Support Vector Machines (SVMs) for learning halfspaces. This result is an application of our general result showing that stable compression schemes [2] of size k have uniformly bounded CMI of order O(k). We further show that an inherent limitation of proper learning of VC classes contradicts the existence of a proper learner with constant CMI, and it implies a negative resolution to an open problem of Steinke and Zakynthinou [3] . We further study the CMI of empirical risk minimizers (ERMs) of class H and show that it is possible to output all consistent classifiers (version space) with bounded CMI if and only if H has a bounded star number [4] . With respect to Item (ii) we prove a general reduction showing that "leave-one-out" analysis is expressible via the CMI framework. As a corollary we investigate the CMI of the one-inclusion-graph algorithm proposed by Haussler et al. [5] . More generally, we show that the CMI framework is universal in the sense that for every consistent algorithm and data distribution, the expected risk vanishes as the number of samples diverges if and only if its evaluated CMI has sublinear growth with the number of samples.