Introduction
Contributions
Backbone | Sparsified | |
---|---|---|
MRE—expected degree | 0.23 | 0.10 |
MRE—expected ego betweenness | 0.29 | 0.17 |
Problem statement
Solution framework
Backboning
Gradient-descent (GD)
Expectation-maximization (EM)
Experiments
Datasets
Brain network
Enron
FriendFeed
Synthetic networks
Dataset | |V| | |E| | \(\overline{{p}}\) | \(\rho\) | \(\overline{D}\) |
---|---|---|---|---|---|
Brain network | 89 | 3916 | 0.526 | 1 | 88 |
Enron | 805 | 3956 | 0.173 | 0.012 | 9.83 |
FriendFeed | 9894 | 172,567 | 0.11 | 0.002 | 34.88 |
ER (high \(\rho\)) | 1000 | 449,550 | 0.49 | 0.9 | 899.1 |
ER (medium \(\rho\)) | 1000 | 249,750 | 0.5 | 0.5 | 499.5 |
ER (low \(\rho\)) | 1000 | 49,950 | 0.53 | 0.1 | 99.9 |
FF (high \(\rho\)) | 1000 | 449,550 | 0.51 | 0.9 | 899.1 |
FF (medium \(\rho\)) | 1000 | 249,750 | 0.49 | 0.5 | 499.5 |
FF (low \(\rho\)) | 1000 | 49,950 | 0.48 | 0.1 | 99.9 |