Graph reduction by local variation
When |
Oct 11, 2018
from 01:30 to 02:30 |
---|---|
Where | room M7.101 |
Attendees |
Andrea Loukas |
Add event to calendar |
vCal iCal |
Can we reduce the size of a graph without significantly altering its basic properties? We will approach the graph reduction problem from the perspective of restricted similarity, a modification of a well-known measure for graph approximation. Our choice is motivated by the observation that restricted similarity implies strong spectral guarantees and can be used to prove statements about certain unsupervised learning problems. The talk will then focus on coarsening, a popular type of graph reduction. We will derive sufficient conditions for a small graph to approximate a larger one in the sense of restricted similarity. Our findings give rise to nearly-linear coarsening algorithms that find coarse graphs of improved quality, often by a large margin, without sacrificing speed.