Algorithm

Type

Time before ESU

Parameter

Deterministic


EDGE GOBNM

EdgeRemoving

O(E)

d

Yes

EDGE BETWEENNESSBNM

EdgeRemoving

O(rEV)

r

Yes

NMFGOBNM

Clustering

O(E(V + l))

d, c, η, β

No

NMFBNM

Clustering

O(EV)

c, η, β

No

VOLTAGEBNM

Clustering

O(E + V)

c

No

 All the algorithms introduced in this paper are compared based on type, time before enumeration, parameter, and whether its deterministic property. Here d is GO depth threshold,l is the number of GO terms associated to the graph G, c is the number of clusters, r is the number of edges to remove, and η, β for sparse NMF computation.