Submodular Maximization beyond Non-negativity: Guarantees, fast algorithms, and applications

Christopher Harshaw, Moran Feldman, Justin Ward, Amin Karbasi

نتاج البحث: فصل من :كتاب / تقرير / مؤتمرمنشور من مؤتمرمراجعة النظراء

ملخص

It is generally believed that submodular functions—and the more general class of 7-weakly submodular functions—may only be optimized under the non-negativity assumption f(S) > 0. In this paper, we show that once the function is expressed as the difference f = g — c, where g is monotone, non-negative, and 7-weakly submodular and c is non-negative modular, then strong approximation guarantees may be obtained. We present an algorithm for maximizing g — c under a fc-cardinality constraint which produces a random feasible set S such that E [g(5)-c(S)≥ (1 - e-e)g(OPT)-c(OPT), whose running time is 0 (2 log2 |), independent of k. We extend these results to the unconstrained setting by describing an algorithm with the same approximation guarantees and faster O(n/c log i/c) runtime. The main techniques underlying our algorithms are two-fold: the use of a surrogate objective which varies the relative importance between g and c throughout the algorithm, and a geometric sweep over possible γ values. Our algorithmic guarantees are complemented by a hardness result showing that no polynomial-time algorithm which accesses g through a value oracle can do better. We empirically demonstrate the success of our algorithms by applying them to experimental design on the Boston Housing dataset and directed vertex cover on the Email EU dataset.

اللغة الأصليةالإنجليزيّة
عنوان منشور المضيف36th International Conference on Machine Learning, ICML 2019
ناشرInternational Machine Learning Society (IMLS)
الصفحات4684-4705
عدد الصفحات22
رقم المعيار الدولي للكتب (الإلكتروني)9781510886988
حالة النشرنُشِر - 2019
الحدث36th International Conference on Machine Learning, ICML 2019 - Long Beach, الولايات المتّحدة
المدة: ٩ يونيو ٢٠١٩١٥ يونيو ٢٠١٩

سلسلة المنشورات

الاسم36th International Conference on Machine Learning, ICML 2019
مستوى الصوت2019-June

!!Conference

!!Conference36th International Conference on Machine Learning, ICML 2019
الدولة/الإقليمالولايات المتّحدة
المدينةLong Beach
المدة٩/٠٦/١٩١٥/٠٦/١٩

ملاحظة ببليوغرافية

Publisher Copyright:
Copyright 2019 by the author(s).

بصمة

أدرس بدقة موضوعات البحث “Submodular Maximization beyond Non-negativity: Guarantees, fast algorithms, and applications'. فهما يشكلان معًا بصمة فريدة.

قم بذكر هذا