PACE 2019: The 4th Iteration Johannes K. Fichte, TU Dresden Markus - - PowerPoint PPT Presentation

▶
pace 2019 the 4th iteration
SMART_READER_LITE
LIVE PREVIEW

PACE 2019: The 4th Iteration Johannes K. Fichte, TU Dresden Markus - - PowerPoint PPT Presentation

Aug 25, 2022 12 likes •832 views

PACE 2019: The 4th Iteration Johannes K. Fichte, TU Dresden Markus Hecher, TU Wien & Univ. of Potsdam IPEC 2019, TU Munich, Germany 1 Johannes K. Fichte & Markus Hecher: PACE 2019 History & Mission of PACE Bart M.P. Jansen 2

slide-1
SLIDE 1

Johannes K. Fichte & Markus Hecher: PACE 2019

PACE 2019: The 4th Iteration

IPEC 2019, TU Munich, Germany

1

Johannes K. Fichte, TU Dresden Markus Hecher, TU Wien & Univ. of Potsdam

slide-2
SLIDE 2

Johannes K. Fichte & Markus Hecher: PACE 2019

History & Mission of PACE

Bart M.P. Jansen

2

slide-3
SLIDE 3

Johannes K. Fichte & Markus Hecher: PACE 2019

History

  • Conceived in fall 2015

parameterized algorithmics should have a greater impact on practice

  • First iteration: 2015/16

○ Track A: Treewidth ○ Track B: Feedback Vertex Set

  • Second iteration: 2016/17

○ Track A: Treewidth ○ Track B: Minimum Fill-In

  • Third iteration: 2017/18

○ Track 1: Steiner tree exact with few terminals ○ Track 2: Steiner tree exact with low treewidth ○ Track 3: Steiner tree heuristic

  • Currently: Fourth iteration

3

slide-4
SLIDE 4

Johannes K. Fichte & Markus Hecher: PACE 2019

Mission

Investigate applicability of algorithmic ideas from parameterized complexity 1. Bridge gap between theory and practice 2. Inspire new theoretical developments 3. Investigate theoretical algorithms in practice 4. Produce accessible implementations & benchmarks 5. Encourage dissemination in scientific papers

4

slide-5
SLIDE 5

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

5

slide-6
SLIDE 6

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

6

slide-7
SLIDE 7

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

7

slide-8
SLIDE 8

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

8

slide-9
SLIDE 9

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

9

slide-10
SLIDE 10

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

10

slide-11
SLIDE 11

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

11

slide-12
SLIDE 12

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

12

slide-13
SLIDE 13

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

13

slide-14
SLIDE 14

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

14

slide-15
SLIDE 15

Johannes K. Fichte & Markus Hecher: PACE 2019

Outcome

  • Previous iterations inspired long list of follow-up works
  • Follow-up Applications and Frameworks
  • Increased awareness of Parameterized Complexity

○ SAT community was particularly interested (this year)

  • Usage of Benchmark instances

○ Almost 12,000 publicly available, citable instances published (this year)

  • Scientific Papers

15

slide-16
SLIDE 16

Johannes K. Fichte & Markus Hecher: PACE 2019

PACE 2019

16

slide-17
SLIDE 17

Johannes K. Fichte & Markus Hecher: PACE 2019

Program and Steering Committee

17

  • Program Committee

Johannes K. Fichte TU Dresden Markus Hecher TU Wien, University of Potsdam Intern: Muhammad A. Dzulfikar University of Indonesia @TU Dresden

  • Steering Committee

Édouard Bonnet Middlesex University Holger Dell IT University of Copenhagen Bart M. P. Jansen Eindhoven University of Technology Thore Husfeldt IT University of Copenhagen and Lund University Petteri Kaski Aalto University Christian Komusiewicz Philipps-Universität Marburg Frances A. Rosamond University of Bergen Florian Sikora LAMSADE, Université Paris Dauphine

slide-18
SLIDE 18

Johannes K. Fichte & Markus Hecher: PACE 2019

  • … for computing resources

We would like to thank our Sponsors...

18

  • … for prizes
slide-19
SLIDE 19

Johannes K. Fichte & Markus Hecher: PACE 2019

Thanks go to

19

  • All the participants of PACE!
  • Intern Muhammad A. Dzulfi

fikar from the University of Indonesia

○ Performing instance selection ○ Support for validating results ○ ….

  • Jan Badura at optil.io, who quickly implemented our requests

○ Using results of several runs for the final results ○ Customizing our judges for optil.io ○ …

slide-20
SLIDE 20

Johannes K. Fichte & Markus Hecher: PACE 2019

Tracks of PACE 2019

20

slide-21
SLIDE 21

Johannes K. Fichte & Markus Hecher: PACE 2019

Track 1: Vertex Cover

  • Among the famous 21 fi

first NP-complete problems by Karp

  • One of the famous, if not the most famous, graph problems
  • Great tradition in parameterized complexity

○ Well studied problem variants ○ Different parameters ○ Kernelizations ○ Applications ○ ...

u
  • Track 1a: Compute a Minimum Vertex Cover (Exact)

21

slide-22
SLIDE 22

Johannes K. Fichte & Markus Hecher: PACE 2019

Instance Selection for Vertex Cover

  • 9,591 instances among 8 different origins

○ PACE 2016 ○ TransitGraphs, Road-graphs ○ SNAP ○ frb ○ ASP Horn backdoors, SAT Horn backdoors ○ SAT2VC

  • Classification by “Difficulty”

(via Gurobi, numVC, Glucose) in intervals →

22

slide-23
SLIDE 23

Johannes K. Fichte & Markus Hecher: PACE 2019

Track 2: Hypertree Decompositions

  • Motivation: Success of PACE 2016 & 2017 (Treewidth)
  • Applications for (Hyper-)tree Decompositions

○ Databases ○ Constraint Programming

  • Track 2a (EXACT): Compute a hypertree decomposition of minimum width
  • Track 2b (HEUR): Heuristically compute a hypertree decomposition of small width

(HEUR)

23

slide-24
SLIDE 24

Johannes K. Fichte & Markus Hecher: PACE 2019

Hypertree Decompositions

  • Given: Hypergraph H (higher arity edges)
  • A Hypertree Decomposition D of H is, roughly speaking,

○ A Tree Decomposition of H ○ + a bag covering function (edge cover) over hyperedges ○ + a certain monotonicity property (Descendent Condition) for the edge cover

  • width(D) is size of the largest edge cover
  • htw(H) smallest width over all Hypertree Decompositions of H

24

slide-25
SLIDE 25

Johannes K. Fichte & Markus Hecher: PACE 2019

Instance Selection for Hypertree Decompositions

  • 2,191 instances from hyperbench, originating from the area of CSP

○ DaimlerChrysler ○ Grid2D ○ MaxSAT, csp_application, csp_random, csp_other ○ CQ

  • Classification of “Difficulty” by means of

○ htdecomp, kdetdecomp ○ Frasmt using the more generalized (fractional / generalized) hypertree decompositions

25

hyperbench.dbai.tuwien.ac.at

slide-26
SLIDE 26

Johannes K. Fichte & Markus Hecher: PACE 2019

PACE 2019: Submission Requirements

26

slide-27
SLIDE 27

Johannes K. Fichte & Markus Hecher: PACE 2019

Submission Requirements

27

1. Solvers + Dependencies have to be open source 2. Source code of solver is maintained on public repository + long term data library 3. A dedicated solver description is required 4. Solvers for Tracks 1a and 2a are provably optimal 1. Submission on optil.io 2. 30 minutes per instance 3. 8 GB RAM per instance

Submission Limits

slide-28
SLIDE 28

Johannes K. Fichte & Markus Hecher: PACE 2019

Participants of PACE 2019

28

slide-29
SLIDE 29

Johannes K. Fichte & Markus Hecher: PACE 2019

Participants

29

  • 18 teams and 33 participants

from 10 countries

slide-30
SLIDE 30

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of PACE 2019

30

slide-31
SLIDE 31

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

31

slide-32
SLIDE 32

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

32

slide-33
SLIDE 33

Johannes K. Fichte & Markus Hecher: PACE 2019 33

slide-34
SLIDE 34

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

34

slide-35
SLIDE 35

Johannes K. Fichte & Markus Hecher: PACE 2019 35

225 €

slide-36
SLIDE 36

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

36

slide-37
SLIDE 37

Johannes K. Fichte & Markus Hecher: PACE 2019 37

250 €

slide-38
SLIDE 38

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

38

slide-39
SLIDE 39

Johannes K. Fichte & Markus Hecher: PACE 2019 39

275 €

slide-40
SLIDE 40

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

40

slide-41
SLIDE 41

Johannes K. Fichte & Markus Hecher: PACE 2019 41

300 €

slide-42
SLIDE 42

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

42

slide-43
SLIDE 43

Johannes K. Fichte & Markus Hecher: PACE 2019 43

450 €

slide-44
SLIDE 44

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

44

slide-45
SLIDE 45

Johannes K. Fichte & Markus Hecher: PACE 2019 45

450 €

slide-46
SLIDE 46

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

47

slide-47
SLIDE 47

Johannes K. Fichte & Markus Hecher: PACE 2019 48

750 €

slide-48
SLIDE 48

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 1: Vertex Cover

  • Track 1a: Minimum Vertex Cover

49

slide-49
SLIDE 49

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 2: Hypertree Decompositions

50

slide-50
SLIDE 50

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 2: Hypertree Decompositions

  • Track 2a: MinHypertreeWidth

51

slide-51
SLIDE 51

Johannes K. Fichte & Markus Hecher: PACE 2019 52

slide-52
SLIDE 52

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 2: Hypertree Decompositions

  • Track 2a: MinHypertreeWidth

54

slide-53
SLIDE 53

Johannes K. Fichte & Markus Hecher: PACE 2019 55

225 €

slide-54
SLIDE 54

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 2: Hypertree Decompositions

  • Track 2a: MinHypertreeWidth

56

slide-55
SLIDE 55

Johannes K. Fichte & Markus Hecher: PACE 2019 57

350 €

slide-56
SLIDE 56

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 2: Hypertree Decompositions

  • Track 2a: MinHypertreeWidth
  • Track 2b: HeurHypertreeWidth

58

slide-57
SLIDE 57

Johannes K. Fichte & Markus Hecher: PACE 2019 59

150 €

slide-58
SLIDE 58

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 2: Hypertree Decompositions

  • Track 2a: MinHypertreeWidth
  • Track 2b: HeurHypertreeWidth

60

slide-59
SLIDE 59

Johannes K. Fichte & Markus Hecher: PACE 2019 61

225 €

slide-60
SLIDE 60

Johannes K. Fichte & Markus Hecher: PACE 2019

Results of Track 2: Hypertree Decompositions

  • Track 2a: MinHypertreeWidth
  • Track 2b: HeurHypertreeWidth

62

slide-61
SLIDE 61

Johannes K. Fichte & Markus Hecher: PACE 2019 63

350 €

slide-62
SLIDE 62

Johannes K. Fichte & Markus Hecher: PACE 2019

Conclusion & Future of PACE

65

slide-63
SLIDE 63

Johannes K. Fichte & Markus Hecher: PACE 2019

Lessons Learned

  • Solver description and public library essential

○ Maybe set solver description (abstract) deadline even at the beginning

  • Selecting instances of moderate difficulty (vertex cover)

○ Possible, but not easy ○ Requires to collect numerous instances

  • Problems and instances should be ready by the end of September

○ Some students lost attention due to late problem announcement

  • Cluster resources essential

○

  • ptil.io troubles during final submission phase

○

  • ptil.io runtime difference was in some cases up to 7 minutes between same instance + solver
  • Score hard to comprehend for submitters

○ limitation of optil.io: normalized scoreboard 0..1 depending

  • n performance of individual solver and instance

66

slide-64
SLIDE 64

Johannes K. Fichte & Markus Hecher: PACE 2019

Future Editions

  • Next years PACE Deadline will be later
  • Process of selecting new PCs ongoing

Problems?

  • Subscribe to the newsletter

and stay tuned!

pacechallenge.org

Hope we see you at PACE 2020.

67

slide-65
SLIDE 65

Johannes K. Fichte & Markus Hecher: PACE 2019

The Winning Team of Track 1a (Vertex Cover)

68

slide-66
SLIDE 66

WeGotYouCovered

IPEC’19 · September 11, 2019 Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash

KIT – University of the State of Baden-Wuerttemberg and National Laboratory of the Helmholtz Association

INSTITUTE OF THEORETICAL INFORMATICS · ALGORITHMICS GROUP

www.kit.edu

The Winning Solver from the PACE 2019 Implementation Challenge, Vertex Cover Track

slide-67
SLIDE 67

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Vertex Cover and Complementary Problems

Input graph Vertex cover Independent Set Clique

1

slide-68
SLIDE 68

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Techniques

Kernelization Iterated Local Search Branch-and-Bound Branch-and-Reduce

· · · 3 7

Branch Reduce Backtrack Reduce (1,2)-Swap

2

slide-69
SLIDE 69

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Kernelization [AI2016]

Reduce Solver Expand Technique from FPT algorithms Applies rich set of reduction rules Significantly reduces graph size

3

slide-70
SLIDE 70

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Iterated Local Search [ARW2012]

(1,2)-Swap Originally developed for independent sets Can often find (near-)optimal solutions Perturbation to escape local optima

4

slide-71
SLIDE 71

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Branch and Reduce [AI2016]

Reduce graph after each branch Additional branching rules to reduce graph size Prune search based on lower bounds Branch Reduce Backtrack

5

slide-72
SLIDE 72

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Branch and Bound [LJM2017]

· · · 3 7

Originally developed for maximum cliques Incremental MaxSAT reasoning to prune search Combination of static and dynamic vertex ordering

6

slide-73
SLIDE 73

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Algorithm Overview

1. Input Graph Kernel Branch-and-Reduce short burst short burst Branch-and-Reduce long run Kernelization 2. Initial Solution Branch-and-Bound long run Branch-and-Bound 6. 3. 4. 5. Iterated Local Search

7

slide-74
SLIDE 74

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Algorithm Overview

1. Input Graph Kernel Branch-and-Reduce short burst short burst Branch-and-Reduce long run Kernelization 2. Initial Solution Branch-and-Bound long run Branch-and-Bound 6. 3. 4. 5. Iterated Local Search

7

slide-75
SLIDE 75

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Algorithm Overview

1. Input Graph Kernel Branch-and-Reduce short burst short burst Branch-and-Reduce long run Kernelization 2. Initial Solution Branch-and-Bound long run Branch-and-Bound 6. 3. 4. 5. Iterated Local Search

7

slide-76
SLIDE 76

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Algorithm Overview

1. Input Graph Kernel Branch-and-Reduce short burst short burst Branch-and-Reduce long run Kernelization 2. Initial Solution Branch-and-Bound long run Branch-and-Bound 6. 3. 4. 5. Iterated Local Search

7

slide-77
SLIDE 77

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Algorithm Overview

1. Input Graph Kernel Branch-and-Reduce short burst short burst Branch-and-Reduce long run Kernelization 2. Initial Solution Branch-and-Bound long run Branch-and-Bound 6. 3. 4. 5. Iterated Local Search

7

slide-78
SLIDE 78

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Algorithm Overview

1. Input Graph Kernel Branch-and-Reduce short burst short burst Branch-and-Reduce long run Kernelization 2. Initial Solution Branch-and-Bound long run Branch-and-Bound 6. 3. 4. 5. Iterated Local Search

7

slide-79
SLIDE 79

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

Instances Solved Over Time

1 10 100 1000 Time t (s) 40 80 120 160 200 Instances solved BnB

  • Kern. + BnB

ILS + BnR BnR WeGotYouCovered

8

slide-80
SLIDE 80

Demian Hespe, Sebastian Lamm, Christian Schulz, Darren Strash – WeGotYouCovered Institute of Theoretical Informatics Algorithmics Group

References

Akiba, Takuya, and Yoichi Iwata. “Branch-and-reduce exponential/FPT algorithms in practice: A case study of vertex cover.” Theoretical Computer Science 609 (2016): 211-225. Li, Chu-Min, Hua Jiang, and Felip Many`

  • a. “On minimization of the

number of branches in branch-and-bound algorithms for the maximum clique problem.” Computers & Operations Research 84 (2017): 1-15. Andrade, Diogo V., Mauricio G. C. Resende, and Renato F . Werneck. “Fast local search for the maximum independent set problem.” Journal

  • f Heuristics 18.4 (2012): 525-547.

Hespe, Demian, Sebastian Lamm, Christian Schulz and Darren Strash “WeGotYouCovered: The Winning Solver from the PACE 2019 Implementation Challenge, Vertex Cover Track.” arXiv preprint arXiv:1908.06795 (2019).

9

slide-81
SLIDE 81

Johannes K. Fichte & Markus Hecher: PACE 2019 69

slide-82
SLIDE 82

Johannes K. Fichte & Markus Hecher: PACE 2019

See you outside at the

70

Thanks again for participating

Poster Session & PACE2020