Significant scales in community structure

Vincent Traag
Vincent TraagResearcher at CWTS
Significant scales in community structure
V.A. Traag1,2, G. Krings3, P. Van Dooren4
1KITLV, Leiden, the Netherlands
2e-Humanities, KNAW, Amsterdam, the Netherlands
3Real Impact, Brussels, Belgium,
4UCL, Louvain-la-Neuve, Belgium
September 17, 2013
eRoyal Netherlands Academy of Arts and Sciences
Humanities
Community Detection
Contant Potts Model (CPM)
• Minimize H(γ) = − ij (Aij − γ)δ(σi , σj )
• Resolution-limit-free
• Internal density pc > γ
• Density between pcd < γ
Community Detection
Contant Potts Model (CPM)
• Minimize H(γ) = − ij (Aij − γ)δ(σi , σj )
• Resolution-limit-free
• Internal density pc > γ
• Density between pcd < γ
Community Detection
Contant Potts Model (CPM)
• Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2
c)
• Resolution-limit-free
• Internal density pc > γ
• Density between pcd < γ
Community Detection
Contant Potts Model (CPM)
• Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2
c)
• Resolution-limit-free
• Internal density pc > γ
• Density between pcd < γ
Community Detection
Contant Potts Model (CPM)
• Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2
c)
• Resolution-limit-free
• Internal density pc > γ
• Density between pcd < γ
Community Detection
Contant Potts Model (CPM)
• Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2
c)
• Resolution-limit-free
• Internal density pc > γ
• Density between pcd < γ
How to choose γ?
Resolution profile
10−3 10−2 10−1 100
103
104
105
106
γ
N E
Significance
How significant is a partition?
Significance
E = 14
E = 9
Fixed partition
E = 11
Better partition
Significance
E = 14
E = 9
Fixed partition
E = 11
Better partition
• Not: Probability to find E edges in partition.
• But: Probability to find partition with E edges.
Subgraph probability
Decompose partition
• Probability to find partition with E edges.
• Probability to find communities with ec edges.
• Asymptotic estimate
• Probability for subgraph of nc nodes with density pc
Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2
cD(pc p)
Significance
• Probability for all communities Pr(σ) ≈
c
exp −n2
cD(pc p) .
• Significance S(σ) = − log Pr(σ) =
c
n2
cD(pc p).
Subgraph probability
Decompose partition
• Probability to find partition with E edges.
• Probability to find communities with ec edges.
• Asymptotic estimate
• Probability for subgraph of nc nodes with density pc
Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2
cD(pc p)
Significance
• Probability for all communities Pr(σ) ≈
c
exp −n2
cD(pc p) .
• Significance S(σ) = − log Pr(σ) =
c
n2
cD(pc p).
Subgraph probability
Decompose partition
• Probability to find partition with E edges.
• Probability to find communities with ec edges.
• Asymptotic estimate
• Probability for subgraph of nc nodes with density pc
Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2
cD(pc p)
Significance
• Probability for all communities Pr(σ) ≈
c
exp −n2
cD(pc p) .
• Significance S(σ) = − log Pr(σ) =
c
n2
cD(pc p).
Subgraph probability
Decompose partition
• Probability to find partition with E edges.
• Probability to find communities with ec edges.
• Asymptotic estimate
• Probability for subgraph of nc nodes with density pc
Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2
cD(pc p)
Significance
• Probability for all communities Pr(σ) ≈
c
exp −n2
cD(pc p) .
• Significance S(σ) = − log Pr(σ) =
c
n2
cD(pc p).
Subgraph probability
Decompose partition
• Probability to find partition with E edges.
• Probability to find communities with ec edges.
• Asymptotic estimate
• Probability for subgraph of nc nodes with density pc
Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2
cD(pc p)
Significance
• Probability for all communities Pr(σ) ≈
c
exp −n2
cD(pc p) .
• Significance S(σ) = − log Pr(σ) =
c
n2
cD(pc p).
Subgraph probability
Decompose partition
• Probability to find partition with E edges.
• Probability to find communities with ec edges.
• Asymptotic estimate
• Probability for subgraph of nc nodes with density pc
Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2
cD(pc p)
Significance
• Probability for all communities Pr(σ) ≈
c
exp −n2
cD(pc p) .
• Significance S(σ) = − log Pr(σ) =
c
n2
cD(pc p).
Significance
10−3 10−2 10−1 100
103
104
105
106
γ
N E
Significance
10−3 10−2 10−1 100
103
104
105
106
γ
N E S
Benchmark
0.25
0.5
0.75
1
NMI
n = 5000, Small
0
1
S
S∗
0 0.2 0.4 0.6 0.8 1
0
1
µ
S∗
S
CPM+Sig
Significance
Modularity
Infomap
OSLOM
Conclusions
• Scan γ efficiently.
• Significance applicable in all methods.
• Correct comparison to random graph.
Traag, Krings, Van Dooren Significant scales in Community Structure
arXiv:1306.3398
Thank you!
Questions?
e-mail: vincent@traag.net twitter: @vtraag
1 of 21

Recommended

Exponential Ranking: Taking into account negative links. by
Exponential Ranking: Taking into account negative links.Exponential Ranking: Taking into account negative links.
Exponential Ranking: Taking into account negative links.Vincent Traag
531 views17 slides
Cooperation, Reputation & Gossiping by
Cooperation, Reputation & GossipingCooperation, Reputation & Gossiping
Cooperation, Reputation & GossipingVincent Traag
563 views25 slides
Cooperation and Reputation by
Cooperation and ReputationCooperation and Reputation
Cooperation and ReputationVincent Traag
443 views41 slides
Social Influence & Community Detection by
Social Influence & Community DetectionSocial Influence & Community Detection
Social Influence & Community DetectionVincent Traag
444 views25 slides
Exponential Ranking: Taking into account negative links. by
Exponential Ranking: Taking into account negative links.Exponential Ranking: Taking into account negative links.
Exponential Ranking: Taking into account negative links.Vincent Traag
318 views18 slides
Dynamics of Discursive Power: The Dutch Debate on Integration. by
Dynamics of Discursive Power: The Dutch Debate on Integration.Dynamics of Discursive Power: The Dutch Debate on Integration.
Dynamics of Discursive Power: The Dutch Debate on Integration.Vincent Traag
385 views17 slides

More Related Content

Viewers also liked

Community Detection with Negative Links by
Community Detection with Negative LinksCommunity Detection with Negative Links
Community Detection with Negative LinksVincent Traag
541 views16 slides
Slice modularity by
Slice modularitySlice modularity
Slice modularityVincent Traag
262 views10 slides
Cooperation, Reputation & Gossiping by
Cooperation, Reputation & GossipingCooperation, Reputation & Gossiping
Cooperation, Reputation & GossipingVincent Traag
374 views26 slides
Cooperation, Reputation & Gossiping by
Cooperation, Reputation & GossipingCooperation, Reputation & Gossiping
Cooperation, Reputation & GossipingVincent Traag
292 views13 slides
Reconstructing Third World Elite Rotation Events from Newspapers by
Reconstructing Third World Elite Rotation Events from NewspapersReconstructing Third World Elite Rotation Events from Newspapers
Reconstructing Third World Elite Rotation Events from NewspapersVincent Traag
8K views16 slides
Social Influence & Popularity by
Social Influence & PopularitySocial Influence & Popularity
Social Influence & PopularityVincent Traag
473 views28 slides

Viewers also liked(20)

Community Detection with Negative Links by Vincent Traag
Community Detection with Negative LinksCommunity Detection with Negative Links
Community Detection with Negative Links
Vincent Traag541 views
Cooperation, Reputation & Gossiping by Vincent Traag
Cooperation, Reputation & GossipingCooperation, Reputation & Gossiping
Cooperation, Reputation & Gossiping
Vincent Traag374 views
Cooperation, Reputation & Gossiping by Vincent Traag
Cooperation, Reputation & GossipingCooperation, Reputation & Gossiping
Cooperation, Reputation & Gossiping
Vincent Traag292 views
Reconstructing Third World Elite Rotation Events from Newspapers by Vincent Traag
Reconstructing Third World Elite Rotation Events from NewspapersReconstructing Third World Elite Rotation Events from Newspapers
Reconstructing Third World Elite Rotation Events from Newspapers
Vincent Traag8K views
Social Influence & Popularity by Vincent Traag
Social Influence & PopularitySocial Influence & Popularity
Social Influence & Popularity
Vincent Traag473 views
Dynamics of Media Attention by Vincent Traag
Dynamics of Media AttentionDynamics of Media Attention
Dynamics of Media Attention
Vincent Traag693 views
Community structure in complex networks by Vincent Traag
Community structure in complex networksCommunity structure in complex networks
Community structure in complex networks
Vincent Traag907 views
Limits of community detection by Vincent Traag
Limits of community detectionLimits of community detection
Limits of community detection
Vincent Traag1.4K views
Public thesis defence: groups and reputation in social networks by Vincent Traag
Public thesis defence: groups and reputation in social networksPublic thesis defence: groups and reputation in social networks
Public thesis defence: groups and reputation in social networks
Vincent Traag607 views
Rapport Projet de Fin d'Etudes by Hosni Mansour
Rapport Projet de Fin d'EtudesRapport Projet de Fin d'Etudes
Rapport Projet de Fin d'Etudes
Hosni Mansour39.6K views
Google Plus et la visibilité: Pourquoi vous devez être sur Google Plus ! by Guinel CADIGNAN
Google Plus et la visibilité: Pourquoi vous devez être sur Google Plus !Google Plus et la visibilité: Pourquoi vous devez être sur Google Plus !
Google Plus et la visibilité: Pourquoi vous devez être sur Google Plus !
Guinel CADIGNAN3.7K views
De l'Internet des Objets à l'Internet des Produits by Renaud Ménérat
De l'Internet des Objets à l'Internet des ProduitsDe l'Internet des Objets à l'Internet des Produits
De l'Internet des Objets à l'Internet des Produits
Renaud Ménérat21.6K views
Droit des cartels et de la concurrence déloyale by fredericborel
Droit des cartels et de la concurrence déloyaleDroit des cartels et de la concurrence déloyale
Droit des cartels et de la concurrence déloyale
fredericborel4.4K views
Renouveler la réflexion et l’action en bibliothèque autour de la notion de bi... by Calimaq S.I.Lex
Renouveler la réflexion et l’action en bibliothèque autour de la notion de bi...Renouveler la réflexion et l’action en bibliothèque autour de la notion de bi...
Renouveler la réflexion et l’action en bibliothèque autour de la notion de bi...
Calimaq S.I.Lex14.1K views

Similar to Significant scales in community structure

Advances in Directed Spanners by
Advances in Directed SpannersAdvances in Directed Spanners
Advances in Directed SpannersGrigory Yaroslavtsev
365 views32 slides
Computational Information Geometry on Matrix Manifolds (ICTP 2013) by
Computational Information Geometry on Matrix Manifolds (ICTP 2013)Computational Information Geometry on Matrix Manifolds (ICTP 2013)
Computational Information Geometry on Matrix Manifolds (ICTP 2013)Frank Nielsen
741 views56 slides
Relaxation methods for the matrix exponential on large networks by
Relaxation methods for the matrix exponential on large networksRelaxation methods for the matrix exponential on large networks
Relaxation methods for the matrix exponential on large networksDavid Gleich
1.4K views36 slides
Participation costs dismiss the advantage of heterogeneous networks in evolut... by
Participation costs dismiss the advantage of heterogeneous networks in evolut...Participation costs dismiss the advantage of heterogeneous networks in evolut...
Participation costs dismiss the advantage of heterogeneous networks in evolut...Naoki Masuda
661 views26 slides
Dotplots for Bioinformatics by
Dotplots for BioinformaticsDotplots for Bioinformatics
Dotplots for Bioinformaticsavrilcoghlan
64.2K views14 slides
Finding similar items in high dimensional spaces locality sensitive hashing by
Finding similar items in high dimensional spaces  locality sensitive hashingFinding similar items in high dimensional spaces  locality sensitive hashing
Finding similar items in high dimensional spaces locality sensitive hashingDmitriy Selivanov
1.2K views20 slides

Similar to Significant scales in community structure(20)

Computational Information Geometry on Matrix Manifolds (ICTP 2013) by Frank Nielsen
Computational Information Geometry on Matrix Manifolds (ICTP 2013)Computational Information Geometry on Matrix Manifolds (ICTP 2013)
Computational Information Geometry on Matrix Manifolds (ICTP 2013)
Frank Nielsen741 views
Relaxation methods for the matrix exponential on large networks by David Gleich
Relaxation methods for the matrix exponential on large networksRelaxation methods for the matrix exponential on large networks
Relaxation methods for the matrix exponential on large networks
David Gleich1.4K views
Participation costs dismiss the advantage of heterogeneous networks in evolut... by Naoki Masuda
Participation costs dismiss the advantage of heterogeneous networks in evolut...Participation costs dismiss the advantage of heterogeneous networks in evolut...
Participation costs dismiss the advantage of heterogeneous networks in evolut...
Naoki Masuda661 views
Dotplots for Bioinformatics by avrilcoghlan
Dotplots for BioinformaticsDotplots for Bioinformatics
Dotplots for Bioinformatics
avrilcoghlan64.2K views
Finding similar items in high dimensional spaces locality sensitive hashing by Dmitriy Selivanov
Finding similar items in high dimensional spaces  locality sensitive hashingFinding similar items in high dimensional spaces  locality sensitive hashing
Finding similar items in high dimensional spaces locality sensitive hashing
Dmitriy Selivanov1.2K views
Дмитрий Селиванов, OK.RU. Finding Similar Items in high-dimensional spaces: L... by Mail.ru Group
Дмитрий Селиванов, OK.RU. Finding Similar Items in high-dimensional spaces: L...Дмитрий Селиванов, OK.RU. Finding Similar Items in high-dimensional spaces: L...
Дмитрий Селиванов, OK.RU. Finding Similar Items in high-dimensional spaces: L...
Mail.ru Group2.1K views
Decomposition and Denoising for moment sequences using convex optimization by Badri Narayan Bhaskar
Decomposition and Denoising for moment sequences using convex optimizationDecomposition and Denoising for moment sequences using convex optimization
Decomposition and Denoising for moment sequences using convex optimization
Chap10 slides by HJ DS
Chap10 slidesChap10 slides
Chap10 slides
HJ DS306 views
Lecture 8: Decision Trees & k-Nearest Neighbors by Marina Santini
Lecture 8: Decision Trees & k-Nearest NeighborsLecture 8: Decision Trees & k-Nearest Neighbors
Lecture 8: Decision Trees & k-Nearest Neighbors
Marina Santini7K views
Csr2011 june15 11_00_sima by CSR2011
Csr2011 june15 11_00_simaCsr2011 june15 11_00_sima
Csr2011 june15 11_00_sima
CSR2011195 views
Return times of random walk on generalized random graphs by Naoki Masuda
Return times of random walk on generalized random graphsReturn times of random walk on generalized random graphs
Return times of random walk on generalized random graphs
Naoki Masuda1K views
Information-theoretic clustering with applications by Frank Nielsen
Information-theoretic clustering  with applicationsInformation-theoretic clustering  with applications
Information-theoretic clustering with applications
Frank Nielsen398 views
On clusteredsteinertree slide-ver 1.1 by VitAnhNguyn94
On clusteredsteinertree slide-ver 1.1On clusteredsteinertree slide-ver 1.1
On clusteredsteinertree slide-ver 1.1
VitAnhNguyn9479 views
Clustering coefficients for correlation networks by Naoki Masuda
Clustering coefficients for correlation networksClustering coefficients for correlation networks
Clustering coefficients for correlation networks
Naoki Masuda83 views
Divide_and_Contrast__Source_free_Domain_Adaptation_via_Adaptive_Contrastive_L... by Huang Po Chun
Divide_and_Contrast__Source_free_Domain_Adaptation_via_Adaptive_Contrastive_L...Divide_and_Contrast__Source_free_Domain_Adaptation_via_Adaptive_Contrastive_L...
Divide_and_Contrast__Source_free_Domain_Adaptation_via_Adaptive_Contrastive_L...
Huang Po Chun5 views
ASCC2022_JunsooKim_220530_.pdf by Junsoo Kim
ASCC2022_JunsooKim_220530_.pdfASCC2022_JunsooKim_220530_.pdf
ASCC2022_JunsooKim_220530_.pdf
Junsoo Kim206 views
Kernel estimation(ref) by Zahra Amini
Kernel estimation(ref)Kernel estimation(ref)
Kernel estimation(ref)
Zahra Amini96 views
pptx - Psuedo Random Generator for Halfspaces by butest
pptx - Psuedo Random Generator for Halfspacespptx - Psuedo Random Generator for Halfspaces
pptx - Psuedo Random Generator for Halfspaces
butest253 views

More from Vincent Traag

Peer review uncertainty at the institutional level by
Peer review uncertainty at the institutional levelPeer review uncertainty at the institutional level
Peer review uncertainty at the institutional levelVincent Traag
233 views12 slides
Replacing peer review by metrics in the UK REF? by
Replacing peer review by metrics in the UK REF?Replacing peer review by metrics in the UK REF?
Replacing peer review by metrics in the UK REF?Vincent Traag
288 views18 slides
Use of the journal impact factor for assessing individual articles need not b... by
Use of the journal impact factor for assessing individual articles need not b...Use of the journal impact factor for assessing individual articles need not b...
Use of the journal impact factor for assessing individual articles need not b...Vincent Traag
161 views13 slides
Uncovering important intermediate publications by
Uncovering important intermediate publicationsUncovering important intermediate publications
Uncovering important intermediate publicationsVincent Traag
182 views19 slides
Complex contagion of campaign donations by
Complex contagion of campaign donationsComplex contagion of campaign donations
Complex contagion of campaign donationsVincent Traag
450 views10 slides
Polarization and consensus in citation networks by
Polarization and consensus in citation networksPolarization and consensus in citation networks
Polarization and consensus in citation networksVincent Traag
801 views50 slides

More from Vincent Traag(9)

Peer review uncertainty at the institutional level by Vincent Traag
Peer review uncertainty at the institutional levelPeer review uncertainty at the institutional level
Peer review uncertainty at the institutional level
Vincent Traag233 views
Replacing peer review by metrics in the UK REF? by Vincent Traag
Replacing peer review by metrics in the UK REF?Replacing peer review by metrics in the UK REF?
Replacing peer review by metrics in the UK REF?
Vincent Traag288 views
Use of the journal impact factor for assessing individual articles need not b... by Vincent Traag
Use of the journal impact factor for assessing individual articles need not b...Use of the journal impact factor for assessing individual articles need not b...
Use of the journal impact factor for assessing individual articles need not b...
Vincent Traag161 views
Uncovering important intermediate publications by Vincent Traag
Uncovering important intermediate publicationsUncovering important intermediate publications
Uncovering important intermediate publications
Vincent Traag182 views
Complex contagion of campaign donations by Vincent Traag
Complex contagion of campaign donationsComplex contagion of campaign donations
Complex contagion of campaign donations
Vincent Traag450 views
Polarization and consensus in citation networks by Vincent Traag
Polarization and consensus in citation networksPolarization and consensus in citation networks
Polarization and consensus in citation networks
Vincent Traag801 views
Structure of media attention by Vincent Traag
Structure of media attentionStructure of media attention
Structure of media attention
Vincent Traag783 views
Dynamical Models Explaining Social Balance by Vincent Traag
Dynamical Models Explaining Social BalanceDynamical Models Explaining Social Balance
Dynamical Models Explaining Social Balance
Vincent Traag270 views
Reputation Dynamics Through Gossiping by Vincent Traag
Reputation Dynamics Through GossipingReputation Dynamics Through Gossiping
Reputation Dynamics Through Gossiping
Vincent Traag385 views

Recently uploaded

1978 NASA News Release Log by
1978 NASA News Release Log1978 NASA News Release Log
1978 NASA News Release Logpurrterminator
11 views146 slides
Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio... by
Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio...Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio...
Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio...Trustlife
127 views17 slides
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ... by
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...ILRI
5 views6 slides
Light Pollution for LVIS students by
Light Pollution for LVIS studentsLight Pollution for LVIS students
Light Pollution for LVIS studentsCWBarthlmew
9 views12 slides
ELECTRON TRANSPORT CHAIN by
ELECTRON TRANSPORT CHAINELECTRON TRANSPORT CHAIN
ELECTRON TRANSPORT CHAINDEEKSHA RANI
10 views16 slides
BLOTTING TECHNIQUES SPECIAL by
BLOTTING TECHNIQUES SPECIALBLOTTING TECHNIQUES SPECIAL
BLOTTING TECHNIQUES SPECIALMuhammadImranMirza2
5 views56 slides

Recently uploaded(20)

Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio... by Trustlife
Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio...Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio...
Discovery of therapeutic agents targeting PKLR for NAFLD using drug repositio...
Trustlife127 views
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ... by ILRI
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...
Small ruminant keepers’ knowledge, attitudes and practices towards peste des ...
ILRI5 views
Light Pollution for LVIS students by CWBarthlmew
Light Pollution for LVIS studentsLight Pollution for LVIS students
Light Pollution for LVIS students
CWBarthlmew9 views
ELECTRON TRANSPORT CHAIN by DEEKSHA RANI
ELECTRON TRANSPORT CHAINELECTRON TRANSPORT CHAIN
ELECTRON TRANSPORT CHAIN
DEEKSHA RANI10 views
RemeOs science and clinical evidence by PetrusViitanen1
RemeOs science and clinical evidenceRemeOs science and clinical evidence
RemeOs science and clinical evidence
PetrusViitanen147 views
Applications of Large Language Models in Materials Discovery and Design by Anubhav Jain
Applications of Large Language Models in Materials Discovery and DesignApplications of Large Language Models in Materials Discovery and Design
Applications of Large Language Models in Materials Discovery and Design
Anubhav Jain13 views
Nitrosamine & NDSRI.pptx by NileshBonde4
Nitrosamine & NDSRI.pptxNitrosamine & NDSRI.pptx
Nitrosamine & NDSRI.pptx
NileshBonde418 views
CSF -SHEEBA.D presentation.pptx by SheebaD7
CSF -SHEEBA.D presentation.pptxCSF -SHEEBA.D presentation.pptx
CSF -SHEEBA.D presentation.pptx
SheebaD715 views
Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe... by Anmol Vishnu Gupta
Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe...Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe...
Study on Drug Drug Interaction Through Prescription Analysis of Type II Diabe...
Conventional and non-conventional methods for improvement of cucurbits.pptx by gandhi976
Conventional and non-conventional methods for improvement of cucurbits.pptxConventional and non-conventional methods for improvement of cucurbits.pptx
Conventional and non-conventional methods for improvement of cucurbits.pptx
gandhi97620 views
Exploring the nature and synchronicity of early cluster formation in the Larg... by Sérgio Sacani
Exploring the nature and synchronicity of early cluster formation in the Larg...Exploring the nature and synchronicity of early cluster formation in the Larg...
Exploring the nature and synchronicity of early cluster formation in the Larg...
Sérgio Sacani910 views
application of genetic engineering 2.pptx by SankSurezz
application of genetic engineering 2.pptxapplication of genetic engineering 2.pptx
application of genetic engineering 2.pptx
SankSurezz14 views
Structure of purines and pyrimidines - Jahnvi arora (11228108), mmdu ,mullana... by jahnviarora989
Structure of purines and pyrimidines - Jahnvi arora (11228108), mmdu ,mullana...Structure of purines and pyrimidines - Jahnvi arora (11228108), mmdu ,mullana...
Structure of purines and pyrimidines - Jahnvi arora (11228108), mmdu ,mullana...
jahnviarora9896 views

Significant scales in community structure

  • 1. Significant scales in community structure V.A. Traag1,2, G. Krings3, P. Van Dooren4 1KITLV, Leiden, the Netherlands 2e-Humanities, KNAW, Amsterdam, the Netherlands 3Real Impact, Brussels, Belgium, 4UCL, Louvain-la-Neuve, Belgium September 17, 2013 eRoyal Netherlands Academy of Arts and Sciences Humanities
  • 2. Community Detection Contant Potts Model (CPM) • Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) • Resolution-limit-free • Internal density pc > γ • Density between pcd < γ
  • 3. Community Detection Contant Potts Model (CPM) • Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) • Resolution-limit-free • Internal density pc > γ • Density between pcd < γ
  • 4. Community Detection Contant Potts Model (CPM) • Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2 c) • Resolution-limit-free • Internal density pc > γ • Density between pcd < γ
  • 5. Community Detection Contant Potts Model (CPM) • Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2 c) • Resolution-limit-free • Internal density pc > γ • Density between pcd < γ
  • 6. Community Detection Contant Potts Model (CPM) • Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2 c) • Resolution-limit-free • Internal density pc > γ • Density between pcd < γ
  • 7. Community Detection Contant Potts Model (CPM) • Minimize H(γ) = − ij (Aij − γ)δ(σi , σj ) = − c(ec − γn2 c) • Resolution-limit-free • Internal density pc > γ • Density between pcd < γ How to choose γ?
  • 8. Resolution profile 10−3 10−2 10−1 100 103 104 105 106 γ N E
  • 10. Significance E = 14 E = 9 Fixed partition E = 11 Better partition
  • 11. Significance E = 14 E = 9 Fixed partition E = 11 Better partition • Not: Probability to find E edges in partition. • But: Probability to find partition with E edges.
  • 12. Subgraph probability Decompose partition • Probability to find partition with E edges. • Probability to find communities with ec edges. • Asymptotic estimate • Probability for subgraph of nc nodes with density pc Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2 cD(pc p) Significance • Probability for all communities Pr(σ) ≈ c exp −n2 cD(pc p) . • Significance S(σ) = − log Pr(σ) = c n2 cD(pc p).
  • 13. Subgraph probability Decompose partition • Probability to find partition with E edges. • Probability to find communities with ec edges. • Asymptotic estimate • Probability for subgraph of nc nodes with density pc Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2 cD(pc p) Significance • Probability for all communities Pr(σ) ≈ c exp −n2 cD(pc p) . • Significance S(σ) = − log Pr(σ) = c n2 cD(pc p).
  • 14. Subgraph probability Decompose partition • Probability to find partition with E edges. • Probability to find communities with ec edges. • Asymptotic estimate • Probability for subgraph of nc nodes with density pc Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2 cD(pc p) Significance • Probability for all communities Pr(σ) ≈ c exp −n2 cD(pc p) . • Significance S(σ) = − log Pr(σ) = c n2 cD(pc p).
  • 15. Subgraph probability Decompose partition • Probability to find partition with E edges. • Probability to find communities with ec edges. • Asymptotic estimate • Probability for subgraph of nc nodes with density pc Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2 cD(pc p) Significance • Probability for all communities Pr(σ) ≈ c exp −n2 cD(pc p) . • Significance S(σ) = − log Pr(σ) = c n2 cD(pc p).
  • 16. Subgraph probability Decompose partition • Probability to find partition with E edges. • Probability to find communities with ec edges. • Asymptotic estimate • Probability for subgraph of nc nodes with density pc Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2 cD(pc p) Significance • Probability for all communities Pr(σ) ≈ c exp −n2 cD(pc p) . • Significance S(σ) = − log Pr(σ) = c n2 cD(pc p).
  • 17. Subgraph probability Decompose partition • Probability to find partition with E edges. • Probability to find communities with ec edges. • Asymptotic estimate • Probability for subgraph of nc nodes with density pc Pr(S(nc, pc) ⊆ G(n, p)) ≈ exp −n2 cD(pc p) Significance • Probability for all communities Pr(σ) ≈ c exp −n2 cD(pc p) . • Significance S(σ) = − log Pr(σ) = c n2 cD(pc p).
  • 18. Significance 10−3 10−2 10−1 100 103 104 105 106 γ N E
  • 19. Significance 10−3 10−2 10−1 100 103 104 105 106 γ N E S
  • 20. Benchmark 0.25 0.5 0.75 1 NMI n = 5000, Small 0 1 S S∗ 0 0.2 0.4 0.6 0.8 1 0 1 µ S∗ S CPM+Sig Significance Modularity Infomap OSLOM
  • 21. Conclusions • Scan γ efficiently. • Significance applicable in all methods. • Correct comparison to random graph. Traag, Krings, Van Dooren Significant scales in Community Structure arXiv:1306.3398 Thank you! Questions? e-mail: vincent@traag.net twitter: @vtraag