The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The...
Transcript of The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The...
![Page 1: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/1.jpg)
The parameterised complexity of subgraphcounting problems
Kitty Meeks
Queen Mary, University of London
Joint work with Mark Jerrum (QMUL)
![Page 2: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/2.jpg)
What is a counting problem?
Decision problems
Given a graph G , does Gcontain a Hamilton cycle?
Given a bipartite graph G ,does G contain a perfectmatching?
![Page 3: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/3.jpg)
What is a counting problem?
Decision problems Counting problems
Given a graph G , does Gcontain a Hamilton cycle?
How many Hamilton cycles arethere in the graph G?
Given a bipartite graph G ,does G contain a perfectmatching?
How many perfect matchingsare there in the bipartite graphG?
![Page 4: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/4.jpg)
What is a parameterised counting problem?
Introduced by Flum and Grohe (2004)
Measure running time in terms of a parameter as well as thetotal input size
Examples:
How many vertex-covers of size k are there in G?How many k-cliques are there in G?Given a graph G of treewidth at most k , how many Hamiltoncycles are there in G?
![Page 5: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/5.jpg)
What is a parameterised counting problem?
Introduced by Flum and Grohe (2004)
Measure running time in terms of a parameter as well as thetotal input size
Examples:
How many vertex-covers of size k are there in G?
How many k-cliques are there in G?Given a graph G of treewidth at most k , how many Hamiltoncycles are there in G?
![Page 6: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/6.jpg)
What is a parameterised counting problem?
Introduced by Flum and Grohe (2004)
Measure running time in terms of a parameter as well as thetotal input size
Examples:
How many vertex-covers of size k are there in G?How many k-cliques are there in G?
Given a graph G of treewidth at most k , how many Hamiltoncycles are there in G?
![Page 7: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/7.jpg)
What is a parameterised counting problem?
Introduced by Flum and Grohe (2004)
Measure running time in terms of a parameter as well as thetotal input size
Examples:
How many vertex-covers of size k are there in G?How many k-cliques are there in G?Given a graph G of treewidth at most k , how many Hamiltoncycles are there in G?
![Page 8: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/8.jpg)
The theory of parameterised counting
Efficient algorithms: Fixed parameter tractable (FPT)Running time f (k) · nO(1)
Intractable problems: #W[1]-hardA #W[1]-complete problem: p-#Clique.
![Page 9: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/9.jpg)
The theory of parameterised counting
Efficient algorithms: Fixed parameter tractable (FPT)Running time f (k) · nO(1)
Intractable problems: #W[1]-hardA #W[1]-complete problem: p-#Clique.
![Page 10: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/10.jpg)
#W[1]-completeness
To show the problem Π′ (with parameter κ′) is #W[1]-hard,we give a reduction from a problem Π (with parameter κ) toΠ′.
An fpt Turing reduction from (Π, κ) to (Π′, κ′) is an algorithmA with an oracle to Π′ such that
1 A computes Π,2 A is an fpt-algorithm with respect to κ, and3 there is a computable function g : N→ N such that for all
oracle queries “Π′(y) =?” posed by A on input x we haveκ′(y) ≤ g(κ(x)).
In this case we write (Π, κ) ≤fptT (Π′, κ′).
![Page 11: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/11.jpg)
#W[1]-completeness
To show the problem Π′ (with parameter κ′) is #W[1]-hard,we give a reduction from a problem Π (with parameter κ) toΠ′.
An fpt Turing reduction from (Π, κ) to (Π′, κ′) is an algorithmA with an oracle to Π′ such that
1 A computes Π,2 A is an fpt-algorithm with respect to κ, and3 there is a computable function g : N→ N such that for all
oracle queries “Π′(y) =?” posed by A on input x we haveκ′(y) ≤ g(κ(x)).
In this case we write (Π, κ) ≤fptT (Π′, κ′).
![Page 12: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/12.jpg)
Subgraph Counting Model
Let Φ be a family (φ1, φ2, . . .) of functions, such that φk is amapping from labelled graphs on k-vertices to {0, 1}.
p-#Induced Subgraph With Property(Φ)Input: A graph G = (V ,E ) and an integer k .Parameter: k.Question: What is the cardinality of the set{(v1, . . . , vk) ∈ V k : φk(G [v1, . . . , vk ]) = 1}?
![Page 13: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/13.jpg)
Subgraph Counting Model
Let Φ be a family (φ1, φ2, . . .) of functions, such that φk is amapping from labelled graphs on k-vertices to {0, 1}.
p-#Induced Subgraph With Property(Φ)Input: A graph G = (V ,E ) and an integer k .Parameter: k.Question: What is the cardinality of the set{(v1, . . . , vk) ∈ V k : φk(G [v1, . . . , vk ]) = 1}?
![Page 14: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/14.jpg)
Examples
p-#Clique
p-#Path
p-#Cycle
p-#Matching
p-#Connected Induced Subgraph
p-#Clique + Independent Set
p-#Planar Induced Subgraph
![Page 15: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/15.jpg)
Examples
p-#Clique
p-#Path
p-#Cycle
p-#Matching
p-#Connected Induced Subgraph
p-#Clique + Independent Set
p-#Planar Induced Subgraph
![Page 16: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/16.jpg)
Examples
p-#Clique
p-#Path
p-#Cycle
p-#Matching
p-#Connected Induced Subgraph
p-#Clique + Independent Set
p-#Planar Induced Subgraph
![Page 17: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/17.jpg)
Examples
p-#Clique
p-#Path
p-#Cycle
p-#Matching
p-#Connected Induced Subgraph
p-#Clique + Independent Set
p-#Planar Induced Subgraph
![Page 18: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/18.jpg)
Examples
p-#Clique
p-#Path
p-#Cycle
p-#Matching
p-#Connected Induced Subgraph
p-#Clique + Independent Set
p-#Planar Induced Subgraph
![Page 19: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/19.jpg)
Examples
p-#Clique
p-#Path
p-#Cycle
p-#Matching
p-#Connected Induced Subgraph
p-#Clique + Independent Set
p-#Planar Induced Subgraph
![Page 20: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/20.jpg)
Complexity Questions
Is the corresponding decision problem in FPT?
Is there a fixed parameter algorithm for p-#InducedSubgraph With Property(Φ)?
Can we approximate p-#Induced Subgraph WithProperty(Φ) efficiently?
![Page 21: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/21.jpg)
Approximation Algorithms
An FPTRAS for a parameterised counting problem Π withparameter k is a randomised approximation scheme that takes aninstance I of Π (with |I | = n), and real numbers ε > 0 and0 < δ < 1, and in time f (k) · g(n, 1/ε, log(1/δ)) (where f is anyfunction, and g is a polynomial in n, 1/ε and log(1/δ)) outputs arational number z such that
P[(1− ε)Π(I ) ≤ z ≤ (1 + ε)Π(I )] ≥ 1− δ.
![Page 22: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/22.jpg)
Problems in our model
DecisionFPT?
FPTRAS?Exact
countingFPT?
p-#Clique N N N
p-#Path Y Y Np-#Cycle
p-#Matching Y Y N
p-#ConnectedInduced Sub-graph
Y Y N
p-#Clique +IndependentSet
Y Y N
Flum & Grohe ’04, Curticapean ’13, Arvind & Raman ’02, Jerrum& M. ’13
![Page 23: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/23.jpg)
The Colourful Version
Suppose the vertices of G are coloured with k colours.
We say a subset of the vertices (or a subgraph) is colourful if itcontains exactly one vertex of each colour.
We define another problem, p-#Multicolour InducedSubgraph with Property(Φ), where we only count colourfullabelled subgraphs satisfying Φ.
![Page 24: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/24.jpg)
Colouring can make problems easier
If the uncoloured version of a parameterised counting problemis in FPT, the multicolour version must also be in FPT: useinclusion-exclusion.
p-#Matching is #W[1]-complete.
p-#Multicolour Matching is in FPT:
There are k!
( k2 )!2
k2
ways to pair up the colours
For each way of pairing up the colours, the number ofmatchings can easily be calculated:
a edges
b edges
c edges
abc matchings
![Page 25: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/25.jpg)
Colouring can make problems easier
If the uncoloured version of a parameterised counting problemis in FPT, the multicolour version must also be in FPT: useinclusion-exclusion.
p-#Matching is #W[1]-complete.
p-#Multicolour Matching is in FPT:
There are k!
( k2 )!2
k2
ways to pair up the colours
For each way of pairing up the colours, the number ofmatchings can easily be calculated:
a edges
b edges
c edges
abc matchings
![Page 26: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/26.jpg)
Colouring can make problems easier
If the uncoloured version of a parameterised counting problemis in FPT, the multicolour version must also be in FPT: useinclusion-exclusion.
p-#Matching is #W[1]-complete.
p-#Multicolour Matching is in FPT:
There are k!
( k2 )!2
k2
ways to pair up the colours
For each way of pairing up the colours, the number ofmatchings can easily be calculated:
a edges
b edges
c edges
abc matchings
![Page 27: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/27.jpg)
Colouring can make problems easier
If the uncoloured version of a parameterised counting problemis in FPT, the multicolour version must also be in FPT: useinclusion-exclusion.
p-#Matching is #W[1]-complete.
p-#Multicolour Matching is in FPT:
There are k!
( k2 )!2
k2
ways to pair up the colours
For each way of pairing up the colours, the number ofmatchings can easily be calculated:
a edges
b edges
c edges
abc matchings
![Page 28: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/28.jpg)
Colouring can make problems harder
p-Clique + Independent Set is in FPT:
By Ramsey, for sufficiently large graphs the answer is always“yes”.
p-Multicolour Clique + Independent Set isW[1]-complete:
Reduction from p-Multicolour Clique.
G
![Page 29: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/29.jpg)
Colouring can make problems harder
p-Clique + Independent Set is in FPT:
By Ramsey, for sufficiently large graphs the answer is always“yes”.
p-Multicolour Clique + Independent Set isW[1]-complete:
Reduction from p-Multicolour Clique.
G
![Page 30: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/30.jpg)
Colouring can make problems harder
p-Clique + Independent Set is in FPT:
By Ramsey, for sufficiently large graphs the answer is always“yes”.
p-Multicolour Clique + Independent Set isW[1]-complete:
Reduction from p-Multicolour Clique.
G
![Page 31: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/31.jpg)
Colouring can make problems harder
p-Clique + Independent Set is in FPT:
By Ramsey, for sufficiently large graphs the answer is always“yes”.
p-Multicolour Clique + Independent Set isW[1]-complete:
Reduction from p-Multicolour Clique.
vG
![Page 32: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/32.jpg)
Hardness I: Properties that hold for few distinct edgedensities
Theorem
Let Φ be a family (φ1, φ2, . . .) of functions φk : {0, 1}(k2) → {0, 1}
that are not identically zero, such that the function mappingk 7→ φk is computable. Suppose that
|{|E (H)| : |V (H)| = k and Φ is true for H}| = o(k2).
Then p-#Induced Subgraph With Property(Φ) is#W[1]-complete.
![Page 33: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/33.jpg)
Hardness I: Properties that hold for few distinct edgedensities
We prove hardness of p-#Multicolour InducedSubgraph with Property(Φ) by means of a reductionfrom p-#Multicolour Clique.
![Page 34: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/34.jpg)
Hardness I: Properties that hold for few distinct edgedensities
We prove hardness of p-#Multicolour InducedSubgraph with Property(Φ) by means of a reductionfrom p-#Multicolour Clique.
![Page 35: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/35.jpg)
Hardness I: Properties that hold for few distinct edgedensities
We prove hardness of p-#Multicolour InducedSubgraph with Property(Φ) by means of a reductionfrom p-#Multicolour Clique.
![Page 36: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/36.jpg)
Hardness I: Properties that hold for few distinct edgedensities
GH
![Page 37: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/37.jpg)
Hardness I: Properties that hold for few distinct edgedensities
GH'
![Page 38: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/38.jpg)
Hardness I: Properties that hold for few distinct edgedensities
Lemma
Let G = (V ,E ) be an n-vertex graph, where n ≥ 2k . Then thenumber of k-vertex subsets U ⊂ V such that U induces either aclique or independent set in G is at least
(2k − k)!
(2k)!
n!
(n − k)!.
![Page 39: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/39.jpg)
Hardness II: Connected subgraphs
Theorem
p-#Connected Induced Subgraph is #W[1]-complete underfpt Turing reductions.
Prove hardness of p-#Multicolour ConnectedInduced Subgraph
Reduction from p-#Multicolour Independent Set
![Page 40: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/40.jpg)
Hardness II: Connected subgraphs
Theorem
p-#Connected Induced Subgraph is #W[1]-complete underfpt Turing reductions.
Prove hardness of p-#Multicolour ConnectedInduced Subgraph
Reduction from p-#Multicolour Independent Set
![Page 41: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/41.jpg)
Hardness II: Connected subgraphs
Associate each colourful set of vertices U with a partition P(U) of{1, . . . , k}.
12 6
53
4
{{1,5,6},{2,3},{4}}
![Page 42: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/42.jpg)
Hardness II: Connected subgraphs
For any partition Pi of {1, . . . , k}, construct GPi.
Suppose Pi = {{1, 2}, {3}, {4}, {5, 6}}:
G 3 4 5 6
7 8 9 10
21
Number of colourful connected induced subgraphs= Number of colourful subsets U ∈ V (G )(k) such thatP(U) ∧ Pi = {{1, . . . , k}}.
![Page 43: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/43.jpg)
Hardness II: Connected subgraphs
For any partition Pi of {1, . . . , k}, construct GPi.
Suppose Pi = {{1, 2}, {3}, {4}, {5, 6}}:
G 3 4 5 6
7 8 9 10
21
Number of colourful connected induced subgraphs= Number of colourful subsets U ∈ V (G )(k) such thatP(U) ∧ Pi = {{1, . . . , k}}.
![Page 44: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/44.jpg)
Hardness II: Connected subgraphs
Let Ni be the number of colourful subsets U ∈ V (k) such thatP(U) = Pi .
Set
aij =
{1 if Pi ∧ Pj = {{1, . . . , k}}0 otherwise.
We can computea1,1 a1,2 · · · a1,Bk
a2,1 a2,2 · · · a2,Bk
......
. . ....
aBk ,1 aBk ,2 · · · aBk ,Bk
·
N0
N1...
NBk
![Page 45: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/45.jpg)
Hardness II: Connected subgraphs
Let Ni be the number of colourful subsets U ∈ V (k) such thatP(U) = Pi .
Set
aij =
{1 if Pi ∧ Pj = {{1, . . . , k}}0 otherwise.
We can computea1,1 a1,2 · · · a1,Bk
a2,1 a2,2 · · · a2,Bk
......
. . ....
aBk ,1 aBk ,2 · · · aBk ,Bk
·
N0
N1...
NBk
![Page 46: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/46.jpg)
Hardness II: Connected subgraphs
Let Ni be the number of colourful subsets U ∈ V (k) such thatP(U) = Pi .
Set
aij =
{1 if Pi ∧ Pj = {{1, . . . , k}}0 otherwise.
We can computea1,1 a1,2 · · · a1,Bk
a2,1 a2,2 · · · a2,Bk
......
. . ....
aBk ,1 aBk ,2 · · · aBk ,Bk
·
N0
N1...
NBk
![Page 47: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/47.jpg)
Approximation algorithm
Theorem
Let Φ = (φ1, φ2, . . .) be a monotone property, and suppose thereexists a positive integer t such that, for each φk , all edge-minimallabelled k-vertex graphs (H, π) such that φk(H) = 1 satisfytreewidth(H) ≤ t. Then there is an FPTRAS for p-#InducedSubgraph With Property(Φ).
![Page 48: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/48.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1
1
1 1 1 1
1 1H G
![Page 49: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/49.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1
1
1 1 1 1
1 1H G
![Page 50: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/50.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
1 1 1 1
1 1H G
![Page 51: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/51.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 1 1 1 1
1 1H G
![Page 52: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/52.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 11 1 1 1
1 1H G
![Page 53: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/53.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 11 14 1 1
1 1H G
![Page 54: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/54.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 11 14 10 1
1 1H G
![Page 55: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/55.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 11 14 10 11
1 1H G
![Page 56: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/56.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 11 14 10 11
0 1 1H G
![Page 57: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/57.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 11 14 10 11
0 16 1H G
![Page 58: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/58.jpg)
Approximation algorithm
Colour the vertices of G with k colours.
For each minimal element H, and each colouring of H with kcolours:
1 1 1 1 1 1
1
1
2 11 14 10 11
0 16 11H G
![Page 59: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/59.jpg)
Open problems
Are there any non-trivial properties in this model that can becounted exactly in FPT time?
Is there an FPTRAS for any monotone property where theminimal elements with the property do not all have boundedtreewidth?
What is the complexity of p-#Induced Subgraph WithProperty(Φ) when φk is true precisely on k-vertex inducedsubgraphs which have an even number of edges?
![Page 60: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/60.jpg)
Open problems
Are there any non-trivial properties in this model that can becounted exactly in FPT time?
Is there an FPTRAS for any monotone property where theminimal elements with the property do not all have boundedtreewidth?
What is the complexity of p-#Induced Subgraph WithProperty(Φ) when φk is true precisely on k-vertex inducedsubgraphs which have an even number of edges?
![Page 61: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/61.jpg)
Open problems
Are there any non-trivial properties in this model that can becounted exactly in FPT time?
Is there an FPTRAS for any monotone property where theminimal elements with the property do not all have boundedtreewidth?
What is the complexity of p-#Induced Subgraph WithProperty(Φ) when φk is true precisely on k-vertex inducedsubgraphs which have an even number of edges?
![Page 62: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/62.jpg)
Open problems
Are there any non-trivial properties in this model that can becounted exactly in FPT time?
Is there an FPTRAS for any monotone property where theminimal elements with the property do not all have boundedtreewidth?
What is the complexity of p-#Induced Subgraph WithProperty(Φ) when φk is true precisely on k-vertex inducedsubgraphs which have an even number of edges?
![Page 63: The parameterised complexity of subgraph counting problemskitty/slides/birmingham-dec13.pdf · The parameterised complexity of subgraph counting problems ... (with parameter ) to](https://reader031.fdocuments.us/reader031/viewer/2022030400/5a73b5107f8b9aa3618b5bf8/html5/thumbnails/63.jpg)
THANK YOU
http://arxiv.org/abs/1308.1575
http://arxiv.org/abs/1310.6524