Experiments using the Fast Eigenspace Approximation using Random Signals (FEARS) method.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data
LICENSE
README.md
approx_filter_jch.m
compressive_kmeans.m
compute_nmi.m
experiment_amazon_clustering.m
experiment_clustering_SBM.m
experiment_eigest_timing.m
experiment_filter_power.m
experiment_various_graphs.m
experiment_viz_large.m
experiment_viz_small.m
print_proj_results.m
rand_index.m
timing_all.m
various_graphs_proj.m

README.md

FEARS-experiments

This repository contains experiments using the Fast Eigenspace Approximation using Random Signals (FEARS) method.

In particular, every experiment presented in the paper FEARS is available as we follow Reproducible Research principles.

If you are looking specifically for the methods to apply FEARS, they are available in the Graph Signal Processing Toolbox from version 0.7.0. The specific functions are gsp_fast_estimate_lk and gsp_eigenspace_approximation.

For this reason, the GSPBox (v0.7.0 and above) is required to run the experiments.