Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Two-Sided Weak Submodularity for Matroid Constrained Optimization and Regression

About

We study the following problem: Given a variable of interest, we would like to find a best linear predictor for it by choosing a subset of $k$ relevant variables obeying a matroid constraint. This problem is a natural generalization of subset selection problems where it is necessary to spread observations amongst multiple different classes. We derive new, strengthened guarantees for this problem by improving the analysis of the residual random greedy algorithm and by developing a novel distorted local-search algorithm. To quantify our approximation guarantees, we refine the definition of weak submodularity by Das and Kempe and introduce the notion of an upper submodularity ratio, which we connect to the minimum $k$-sparse eigenvalue of the covariance matrix. More generally, we look at the problem of maximizing a set function $f$ with lower and upper submodularity ratio $\gamma$ and $\beta$ under a matroid constraint. For this problem, our algorithms have asymptotic approximation guarantee $1/2$ and $1-e^{-1}$ as the function is closer to being submodular. As a second application, we show that the Bayesian A-optimal design objective falls into our framework, leading to new guarantees for this problem as well.

Theophile Thiery, Justin Ward• 2021

Related benchmarks

TaskDatasetResultRank
Coverage MaximizationVideo Summarization (n=20, k=5)
Coverage Score34
6
Coverage MaximizationVideo Summarization n=40 k=8
Objective Score71
6
Coverage MaximizationVideo Summarization n=50, k=10
Coverage Score89
6
Bayesian A-optimal DesignSONAR
Objective Value610.4
6
Video SummarizationAnimation website Video sequence V2 1.0
Object Score79.95
6
Video SummarizationVSUMM TV Show V8 1.0
Obj Score40.7
6
Bayesian A-optimal DesignEunite 2001
Objective Value102.4
6
Coverage MaximizationVideo Summarization n=30, k=6
Coverage Objective Score34
6
Video SummarizationCooking website Video sequence 1.0
Object Score50.56
6
Video SummarizationVSUMM Soccer V3 1.0
Object Score50.2
6
Showing 10 of 18 rows

Other info

Follow for update