Code for approximate graph reduction techniques for cardinality-based DSFM, from paper

"Approximate Decomposable Submodular Function Minimization for Cardinality-Based Components"

Nate Veldt, Austin Benson, Jon Kleinberg. NeurIPS 2021.

The include folder contains implementations for outside code needed for experimental comparisons.

The src folder includes implementations of our main methods.

There is a folder for each of the main experiments (one for image segmentation, one for hypergraph clustering).

Code for running image segmentation experiments with competing continuous optimization techniques is given in

include/DSFM-with-incidence-relations-v2

In order to reproduce these experiments, see

Run_all_image_exps.m

You will need to place the stackoverflow-answers dataset in the data folder in order to run experiments

https://www.cs.cornell.edu/~arb/data/stackoverflow-answers/

To reproduce experiments, run

stackoverflow_runall.jl

This will take a long time, as this file runs 4500 individual local clustering experiments.