Transcript LP-Based Parameterized Algorithms for Separation
LP-Based Parameterized Algorithms for Separation Problems D. Lokshtanov , N.S. Narayanaswamy V. Raman, M.S. Ramanujan S. Saurabh
Message of this talk
It was open for quite a while whether Odd Cycle Transversal and Almost 2-Sat are FPT .
A simple branching algorithm for Vertex Cover known since the mid 90βs solves both problems in time π β (4 π ) .
Some more work gives π β (2.32
π ) .
Results
( Above LP ) Multiway Cut 4 πβπΏπ
[CPPW11]
2.32
πβπΏπ ( Above LP ) Vertex Cover Almost 2 -SAT 2.32
π 2.32
π Odd Cycle Transversal
How does one get a 4 k-LP algorithm?
Branching: on both sides k-LP least Β½ .
decreases by at How to improve ? Decrease k-LP more.
Multiway Cut
In:
Graph G , set T of vertices, integer k .
Question:
of G\S βπ β π πΊ such that no component has at least two vertices of T ?
FPT by Marx , 04 Faster FPT by Chen et al, 07 Fastest FPT and FPT/ k-LP by Cygan et al, 11
Vertex Cover
In:
G , t
Question:
edge in G βπ β π πΊ , π β€ t such that every has an endpoint in S ?
Long story...
Here: π β (2.32
π‘βπΏπ ) .
Almost 2-SAT
In:
2 SAT formula π , integer k
Question:
Can we remove k variables from π and make it satisfiable?
FPT by Razgon and OβSullivan , 08 Here: π β (2.32
π ) .
Odd Cycle Transversal
In:
G , k
Question:
bipartite?
βπ β π πΊ , π β€ k such that G\S is FPT: π β (3 π ) by Reed et al.
Here: π β (2.32
π ) .
Vertex Cover
In:
G , t
Question:
edge in G βπ β π πΊ , π β€ t such that every has an endpoint in S ?
Minimize βπ₯π βπ’π£ β πΈ πΊ : π₯π’ + π₯π£ β₯ 1 π₯ π β₯ 0 π₯ π β
Z
Vertex Cover Above LP
In:
G , t
Question:
edge in G βπ β π πΊ , π β€ t such that every has an endpoint in S ?
Running Time:
π β (π π‘ β πΏπ ) , where LP value of the optimum LP solution.
is the π = π‘ β πΏπ
x z Odd Cycle Transversal ο Almost 2-Sat y π₯ β¨ Β¬π¦ Β¬π₯ β¨ π¦ x π₯ β¨ Β¬π§ Β¬π₯ β¨ π§ π¦ β¨ Β¬π§ z y Β¬π¦ β¨ π§
Almost 2-SAT ο
Vertex Cover/
t-LP π₯ π₯ β¨ π¦ π¦ π¦ β¨ Β¬π§ π§ Β¬π₯ Β¬π¦ Β¬π§
Nemhauser Trotter Theorem
(a) There is always an optimal solution to Vertex 1 Cover LP that sets variables to {0 , , 1} .
2 1 (b) For any {0 , , 1} βsolution there is a matching 2 from the 1 -vertices to the 0 -vertices, saturating the 1 -vertices.
Nemhauser Trotter Proof
+ Ο΅ + Ο΅ Ο΅ Ο΅ Ο΅ < π π > π π π π
Reduction Rule
If exists optimal LP solution that sets x v to 1 , then exists optimal vertex cover that selects v .
ο Remove v from G and decrease t by 1 .
Correctness follows from Nemhauser Trotter Polynomial time by LP solving.
Branching
Pick an edge uv . Solve (G\u, t-1) and (G\v, t-1) .
LP(G\u) > LP(G) β 1 since otherwise there is an optimal LP solution for G that sets u to 1 .
Then LP(G\u) β₯ LP(G) β 1 2
Branching - Analysis
LP β t drops by Β½ ... in both branches!
π π β€ 2π π β 2 β€ 4 π Total time: π β (4 π‘βπΏπ )
Caveat: The reduction does not increase the measure!
Moral
Nemhauser Trotter
reduction + classic Β«
branch on an edge
Β» gives π β 4 πΏπβπ‘ time algorithm for Vertex Cover and π β 4 π time algorithm for Odd Cycle Transversal and Almost 2-Sat .
Can we do better?
Surplus
The
surplus
be negative!
of a set I is |N(I)| β |I| . The surplus can 1 In any {0 , , 1} -LP solution, the total weight is 2 n/2 + surplus(V 0 )/2 .
Solving the Vertex Cover LP an independent set I is equivalent to finding of minimum surplus.
Surplus and Reductions
If Β« all Β½ Β» is the unique LP optimum then surplus(I) > 0 for all independent sets.
Can we say anything meaningful for independent sets of surplus 1? 2? k?
Surplus Branching Lemma
Let I be an independent set in G with minimum surplus . There exists an optimal vertex cover C that either contains I or avoids I .
Surplus Branching Lemma Proof
πΆ β© πΌ πΆ\πΌ I N(I) R
Branching Rule
Find an independent set Solve (G\I, t-|I|) I of minimum surplus. and (G\N(I), t-|N(I)|) .
LP(G\I) > LP(G) - |I| , since otherwise LP(G) optimal solution that sets I to 1 .
has an So πΏπ πΊ\I β₯ πΏπ πΊ β πΌ + 1 2 t-LP drops by at least Β½ .
Branching Rule Analysis Contβd
Analyzing the (G\N(I), t-N(I)) side: LP(G\N(I)) + |N(I)| = n/2 + surplus(I)/2 β₯ LP G + surplus(I)/2 So LP(G\N(I)) β₯ LP(G) β |N(I)| + surplus(I)/2 t-LP drops by at least surplus(I)/2
Branching Summary
The measure k-LP drops by (Β½, surplus(I)/2) .
Will see that independent sets of surplus 1 be reduced in polynomial time !
can Measure drops by (Β½,1) giving a π β time algorithm for Vertex Cover 2.618
π‘βπΏπ
Reducing Surplus 1 sets.
Lemma:
and N(I) If surplus(I) = 1 , I has minimum surplus is not independent then there exists an optimum vertex cover containing N(I) .
I N(I) R
Reducing Surplus 1 sets.
Reduction Rule:
surplus and N(I) If surplus(I) = 1 , I has minimum is independent then solve (Gβ,t-|I|) where single vertex v .
Gβ is G with N[I] contracted to a I N(I) R
Summary
Nemhauser Trotter lets us assume surplus > 0 More rules let us assume surplus > 1 ( β₯ 2 ) * If surplus π β 2.618
β₯ 2 π‘βπΏπ then branching yields time for Vertex Cover The correctness of these rules were also proved by NT!
Can we do better?
Can get down to π β (2.32
π‘βπΏπ ) by more clever branching rules. Yields π β (2.32
π‘βπΏπ ) for Almost 2-SAT and Odd Cycle Transversal .
Should not be the end of the story.
Better OCT?
Can we get down to π β (2 π ) Transversal ?
for Odd Cycle
LP Branching in other cases
I believe many more problems should have FPT algorithms by LP -guided branching.
What about ... ( Directed ) Feedback Vertex Set , parameterized by solution size k ?