Differentiable Greedy Submodular Maximization: Guarantees, Gradient Estimators, and Applications

05/06/2020
by   Shinsaku Sakaue, et al.
0

We consider making outputs of the greedy algorithm for monotone submodular function maximization differentiable w.r.t. parameters of objective functions; this is motivated by many applications, e.g., sensitivity analysis and end-to-end learning. Our contribution is a theoretically guaranteed and widely applicable smoothing framework based on randomization. We prove that our smoothed greedy algorithm almost recovers original approximation guarantees in expectation for the cases of cardinality and κ-extensible system constrains. We also show how to efficiently compute unbiased gradient estimators of any expected output-dependent quantities by sampling outputs. We demonstrate the utility and effectiveness of our framework by applying it to various situations including the aforementioned ones.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

page 4

05/06/2020

Differentiable Greedy Submodular Maximization with Guarantees and Gradient Estimators

We consider making outputs of the greedy algorithm for monotone submodul...
08/17/2019

Approximation Guarantees of Stochastic Greedy Algorithms for Non-monotone Submodular Maximization with a Size Constraint

The stochastic greedy algorithm (SG) is a randomized version of the gree...
03/05/2018

Differentiable Submodular Maximization

We consider learning of submodular functions from data. These functions ...
04/30/2020

Weakly Submodular Function Maximization Using Local Submodularity Ratio

Weak submodularity is a natural relaxation of the diminishing return pro...
11/02/2020

Unified greedy approximability beyond submodular maximization

We consider classes of objective functions of cardinality constrained ma...
10/22/2020

Planning with Submodular Objective Functions

We study planning with submodular objective functions, where instead of ...
06/03/2020

Batch greedy maximization of non-submodular functions: Guarantees and applications to experimental design

We propose and analyze batch greedy heuristics for cardinality constrain...
This week in AI

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