E N D
1. ApproximateMax-integral-flow/min-cut Theorems Kenji Obata
UC Berkeley
June 15, 2004
2. Multicommodity Flow Graph G, edge capacities c, demands K
3. Multicommodity Flow K-partition
4. Multicommodity Flow K-cut
5. Multicommodity Flow
6. Multicommodity Flow
7. Multicommodity Flow Interesting because flow problem is poly computable, while cut problem is NP-hard => approximation algorithms for cut problemsInteresting because flow problem is poly computable, while cut problem is NP-hard => approximation algorithms for cut problems
8. Integral Multicommodity Flow Suppose c is integral. Can we find integral f ?
for one commodity, yes [Ford-Fulkerson]
in general, no [Garg]
Both flow [GVY] and cut [DJPSY] problems are NP-hard Beautiful property of Ford-Fulkerson Thm is integral capacities => integral flow.
Unlike fractional case, no analogous equality nor approximation.
Also, *both* flow and cut problems are NP-hard.Beautiful property of Ford-Fulkerson Thm is integral capacities => integral flow.
Unlike fractional case, no analogous equality nor approximation.
Also, *both* flow and cut problems are NP-hard.
9. Integral Multicommodity Flow Takeaway: This turns out to be an extremely natural parametrization of multicommodity flow problems.Takeaway: This turns out to be an extremely natural parametrization of multicommodity flow problems.
10. Integral Multicommodity Flow Dense graph has unit capacities.Dense graph has unit capacities.
11. Integral Multicommodity Flow Algorithmic:
Construct an integral flow
or a proof that the K-cut condition is violated
=> edge-disjoint path problems
=> odd circuit cover problems
=> property testing In other words, when multicut is dense, not only does min-cut become a good approximation to max-integral-flow, but it becomes easy to do integral routing between terminals.In other words, when multicut is dense, not only does min-cut become a good approximation to max-integral-flow, but it becomes easy to do integral routing between terminals.
12. Algorithm (general graphs)
13. Algorithm (general graphs)
15. Constructing g(t) Radius refers to # edges (not weighted)
G some family of graphsRadius refers to # edges (not weighted)
G some family of graphs
16. Constructing g(t)
17. Constructing g(t)
18. Bounding f(r) General graphs
Reinterpret [GVY] applied to original graph metric
(Note: Makes no sense)
Planar graphs
[Klein-Plotkin-Rao]
Dense graphs
19. Bounding f(r) (dense case) |E(G)| >= dn2, d > 0, c {0,1}E
B(v, r) = ball of radius r around v, boundary Bo(v, r)
20. Choose arbitrary vertex v, set r = 0
While |Bo(v, r)| |Bo(v, r+1)| > a |B(v, )| |B(v, r)|, grow Bounding f(r) (dense case)
21. Bounding f(r) (dense case) Each ball has low radius
Proof:
22. Bounding f(r) (dense case) Induced multicut has low density
Proof:
23. Proof of Theorem Suppose every K-cut has weight >= eC
Claim: $ K-path of length <= g(e):
24. Proof of Theorem
25. Proof of Theorem (contd) Delete path p (|p| <= g(e)) and iterate
c = c p ; e = e p/C
Witness for flow f, residual multicut m
26. Edge-disjoint paths Corollary:
If G has degree bound D, min-multicut em then
27. Motivation (Property Testing) Given bounded degree graph G
Want to distinguish whether
G has a certain property
or is far (en entries) from having the property
In sub-linear (constant?) time
Example: Coloring problems
No sub-linear algorithms for 3-coloring [BOT]
2-coloring has complexity ~O(n1/2)
28. Testing 2-Colorability Fix max-cut
Set G = {crossing edges}, K = {internal edges}
=> min-multicut has weight >= em
29. Testing 2-Colorability (planar case)
30. Thank you