Nearly Linear-Time, Parallelizable Algorithms for Non-Monotone Submodular Maximization

09/03/2020 ∙ by Alan Kuhnle, et al. ∙ 0

We study parallelizable algorithms for maximization of a submodular function, not necessarily monotone, with respect to a cardinality constraint k. We improve the best approximation factor achieved by an algorithm that has optimal adaptivity and query complexity, up to logarithmic factors in the size n of the ground set, from 0.039 - ϵ to 0.193 - ϵ. We provide two algorithms; the first has approximation ratio 1/6 - ϵ, adaptivity O( log n ), and query complexity O( n log k ), while the second has approximation ratio 0.193 - ϵ, adaptivity O( log^2 n ), and query complexity O(n log k). Heuristic versions of our algorithms are empirically validated to use a low number of adaptive rounds and total queries while obtaining solutions with high objective value in comparison with highly adaptive approximation algorithms.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

page 4

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.