When a given weighted directed acyclic graph represents some object flow, what is the most efficient algorithm to find the equivalent graph sum of which edges is minimum? Here, 'equivalent' means the end state of the objects is the same.
The picture below is the trivial example of this problem.
Above, 'simplest' means least weight sum.
updated
Sorry, my problem specification was ambiguous. I updated the sample problem explanation.
update 2
To clarify the specification, I provide mathematical formulation.
Definition 1 transport mass map
Transport mass map T
is a map between directed weighted graph to a set of pairs of node and transported volume.
for example,
Definition 2 equivalence in terms of transportation
In this problem, two graphs G1
and G2
are equivalent iff T(G1)
is T(G2)
the same sets.
Lemma 1 equivalence relation
If G1 ~ G2
means G1
is equivalent with G2
in this sense, ~
is equivalence relation in a directed weighted graph set.
Definition 3 equivalence set
For graph G
, S*(G)
is a set of graph such that
if a graph g
in S*(G)
, g~G
, and vice versa.
Definition 3 minimal graph in terms of transportation
Graph G-
is called minimal for G
iff the sum of G-
's node weights is minimum of all graph in S*(G)
Problem statement 1
What is (the most) efficient method to find G-
when G
is given.
Problem statement 2 (added)
What is the method above called generally?