We consider the problem of maximizing a submodular function when given access to its approximate version. Submodular functions are heavily studied in a wide variety of disciplines since they are used to model many real world phenomena and are amenable to optimization. There are many cases however in which the phenomena we observe is only approximately submodular and the optimization guarantees cease to hold. In this paper we describe a technique that yields strong guarantees for maximization of monotone submodular functions from approximate surrogates under cardinality and intersection of matroid constraints. In particular, we show tight guarantees for maximization under a cardinality constraint and 1/(1 + P) approximation under intersection of P matroids.
|Number of pages
|Advances in Neural Information Processing Systems
|Published - 2018
|32nd Conference on Neural Information Processing Systems, NeurIPS 2018 - Montreal, Canada
Duration: 2 Dec 2018 → 8 Dec 2018
Bibliographical notePublisher Copyright:
© 2018 Curran Associates Inc..All rights reserved.