SlideShare a Scribd company logo
1 of 33
Download to read offline
Overlapping Community detection using Bayesian Non
Negative Matrix Factorization
Rajkumar Singh
Rishi Barua
Guide: Dr. Ashish Anand
Dept. of Computer Science
Indian Institute of Technology, Guwahati
April 18, 2013
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 1 / 15
Network Paradigm
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
Network Paradigm
V ∈ RN×N is the adjacency matrix
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
Network Paradigm
V ∈ RN×N is the adjacency matrix
vij can be boolean or denote connection weight
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
Network Paradigm
V ∈ RN×N is the adjacency matrix
vij can be boolean or denote connection weight
ki = j vij degree of node i
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
Network Paradigm
V ∈ RN×N is the adjacency matrix
vij can be boolean or denote connection weight
ki = j vij degree of node i
data in relational form
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
Commnity Detection
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Commnity Detection
Q. What is Community ?
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Commnity Detection
Q. What is Community ?
As such no Specific Definition. It is Context Dependent.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Commnity Detection
Q. What is Community ?
As such no Specific Definition. It is Context Dependent.
Defn. : Here we define communities as Modules which are subgraphs with
more links connecting the nodes inside than outside them.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Commnity Detection
Q. What is Community ?
As such no Specific Definition. It is Context Dependent.
Defn. : Here we define communities as Modules which are subgraphs with
more links connecting the nodes inside than outside them.
A given real world network is assumed to be clustered into a number
of latent classes of nodes.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Commnity Detection
Q. What is Community ?
As such no Specific Definition. It is Context Dependent.
Defn. : Here we define communities as Modules which are subgraphs with
more links connecting the nodes inside than outside them.
A given real world network is assumed to be clustered into a number
of latent classes of nodes.
These nodes form regions of increased connectivity in the network.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Commnity Detection
Q. What is Community ?
As such no Specific Definition. It is Context Dependent.
Defn. : Here we define communities as Modules which are subgraphs with
more links connecting the nodes inside than outside them.
A given real world network is assumed to be clustered into a number
of latent classes of nodes.
These nodes form regions of increased connectivity in the network.
These communities usually reflect functional modules that affect the
overall behavior of the system.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Commnity Detection
Q. What is Community ?
As such no Specific Definition. It is Context Dependent.
Defn. : Here we define communities as Modules which are subgraphs with
more links connecting the nodes inside than outside them.
A given real world network is assumed to be clustered into a number
of latent classes of nodes.
These nodes form regions of increased connectivity in the network.
These communities usually reflect functional modules that affect the
overall behavior of the system.
Examples: Friend cliques in social networks, Similar proteins in a
protein interaction network, Research groups in a scientific
collaboration network
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
Community Detection
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 4 / 15
Non-negative Matrix Factorzation
Decompose data matrix V to a product of two other matrices W , H under
non-negativity constraints.
V ∼ ˆV = WH V , ˆV ∈ RF×N and W ∈ RF×K , H ∈ RK×N so that,
FK + KN < FN
Non-negativity constraints avoid the problem of an ill-posed solution
They also reflect the idea of parts-based representation. V can be
expressed as additive combination of certain basis structures defined
by wk, given an encoding hk
ˆV =
k
w:khk: (1)
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 5 / 15
Poisson Model
Given an Adjacency matrix V ∈ RN×N
Assume, pairwise interactions vij is generated by a Poisson
distribution with rate ˆvij .
Hence ˆV ∼ V
Expectation network ˆV is composed of W and H, so that ˆV = WH
Inner rank K : Unknown number of communities
wik, hkj ∈ 1, . . . , K: The contribution of each latent community to
ˆvij = k wikhkj
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 6 / 15
Cost Function
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
Cost Function
posterior:
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
Cost Function
posterior:
p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2)
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
Cost Function
posterior:
p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2)
by taking the − log of: p(V , W , H)p(W |β)p(H|β)p(β),
we define cost as U.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
Cost Function
posterior:
p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2)
by taking the − log of: p(V , W , H)p(W |β)p(H|β)p(β),
we define cost as U.
U = − log p(V |W , H) − log p(W |β) − log p(W |β) − log p(β) (3)
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
Cost Function
posterior:
p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2)
by taking the − log of: p(V , W , H)p(W |β)p(H|β)p(β),
we define cost as U.
U = − log p(V |W , H) − log p(W |β) − log p(W |β) − log p(β) (3)
U is our minimization objective under non-negativity constraints.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
Parameter inference
U = i j vij log
vij
ˆvij
+ ˆvij
+
1
2 k i βkw2
ik + j βkh2
kj − 2N log βk
+ k βkbk − (ak − 1) log βk + k
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 8 / 15
Parameter inference
U = i j vij log
vij
ˆvij
+ ˆvij
+
1
2 k i βkw2
ik + j βkh2
kj − 2N log βk
+ k βkbk − (ak − 1) log βk + k
We find w∗, h∗,β∗ that minimize U using the gradient descent
algorithm.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 8 / 15
Parameter inference
U = i j vij log
vij
ˆvij
+ ˆvij
+
1
2 k i βkw2
ik + j βkh2
kj − 2N log βk
+ k βkbk − (ak − 1) log βk + k
We find w∗, h∗,β∗ that minimize U using the gradient descent
algorithm.
The effective number of communities K is the number of non-zero
columns of W∗ and H∗
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 8 / 15
Results
W∗, H∗ describe a bipartite network of node allocations to
communities.
If our original adjacency matrix V is symmetric, then W∗ = HT
∗ .
Each wik or hki denotes the participation strength of node i to
community k.
The ith row of W or column of H describes a soft-membership
distribution of node i across communities.
Varying node participation scores allow us to describe overlaps
between communities in a disciplined manner.
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 9 / 15
Example: 1
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 10 / 15
Example: 1
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 11 / 15
Example: 1
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 12 / 15
Example: 2
We now take a 1000X1000 Adjacency matrix V , which is very sparse.
The graph is a non-community graph, i.e. having no community
structure.
Several of the nodes are isolated.
Solutions from EO, Louvian (as described in the paper) offer higher
modularity than NMF.
NMF clearly shows that the modularity is low and the nodes have no
preference to lie in any particular community, or are non-communal.
NMF does not suffer from resolution limit of modularity (which
groups smaller communities), as can be clearly seen.
The Output shows 1 community, with 428 isolated nodes i.e.
Non-communal
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 13 / 15
Example: 2
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 14 / 15
References
1 Overlapping community detection using Bayesian non-negative matrix
factorization - I. Psorakis, S. Roberts and M.Ebden
2 Signal Processing with Adaptive Sparse Structured Representations -
V.Tan and C.Fevotte
3 D.D. Lee and H.S. Seung, Nature ,401
R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 15 / 15

More Related Content

Viewers also liked

Community Detection with Networkx
Community Detection with NetworkxCommunity Detection with Networkx
Community Detection with NetworkxErika Fille Legara
 
Big data matrix factorizations and Overlapping community detection in graphs
Big data matrix factorizations and Overlapping community detection in graphsBig data matrix factorizations and Overlapping community detection in graphs
Big data matrix factorizations and Overlapping community detection in graphsDavid Gleich
 
Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...
Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...
Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...Spark Summit
 
Community detection in graphs
Community detection in graphsCommunity detection in graphs
Community detection in graphsNicola Barbieri
 
Applications of community detection in bibliometric network analysis
Applications of community detection in bibliometric network analysisApplications of community detection in bibliometric network analysis
Applications of community detection in bibliometric network analysisNees Jan van Eck
 
Community Detection in Social Media
Community Detection in Social MediaCommunity Detection in Social Media
Community Detection in Social MediaSymeon Papadopoulos
 

Viewers also liked (6)

Community Detection with Networkx
Community Detection with NetworkxCommunity Detection with Networkx
Community Detection with Networkx
 
Big data matrix factorizations and Overlapping community detection in graphs
Big data matrix factorizations and Overlapping community detection in graphsBig data matrix factorizations and Overlapping community detection in graphs
Big data matrix factorizations and Overlapping community detection in graphs
 
Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...
Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...
Dynamic Community Detection for Large-scale e-Commerce data with Spark Stream...
 
Community detection in graphs
Community detection in graphsCommunity detection in graphs
Community detection in graphs
 
Applications of community detection in bibliometric network analysis
Applications of community detection in bibliometric network analysisApplications of community detection in bibliometric network analysis
Applications of community detection in bibliometric network analysis
 
Community Detection in Social Media
Community Detection in Social MediaCommunity Detection in Social Media
Community Detection in Social Media
 

Similar to Overlapping community Detection Using Bayesian NMF

Community DetectionSlide
Community DetectionSlideCommunity DetectionSlide
Community DetectionSlideAshwini Tokekar
 
IRJET- Link Prediction in Social Networks
IRJET- Link Prediction in Social NetworksIRJET- Link Prediction in Social Networks
IRJET- Link Prediction in Social NetworksIRJET Journal
 
Action and content based Community Detection in Social Networks
Action and content based Community Detection in Social NetworksAction and content based Community Detection in Social Networks
Action and content based Community Detection in Social Networksritesh_11
 
Clustering large probabilistic graphs
Clustering large probabilistic graphsClustering large probabilistic graphs
Clustering large probabilistic graphsecway
 
User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...
User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...
User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...IIIT Hyderabad
 

Similar to Overlapping community Detection Using Bayesian NMF (7)

Community DetectionSlide
Community DetectionSlideCommunity DetectionSlide
Community DetectionSlide
 
IRJET- Link Prediction in Social Networks
IRJET- Link Prediction in Social NetworksIRJET- Link Prediction in Social Networks
IRJET- Link Prediction in Social Networks
 
Action and content based Community Detection in Social Networks
Action and content based Community Detection in Social NetworksAction and content based Community Detection in Social Networks
Action and content based Community Detection in Social Networks
 
Community detection
Community detectionCommunity detection
Community detection
 
Clustering large probabilistic graphs
Clustering large probabilistic graphsClustering large probabilistic graphs
Clustering large probabilistic graphs
 
User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...
User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...
User Identity Linkage: Data Collection, DataSet Biases, Method, Control and A...
 
Pula 5 Giugno 2007
Pula 5 Giugno 2007Pula 5 Giugno 2007
Pula 5 Giugno 2007
 

More from Raj Sikarwar

Rabindranath tagore Biography
Rabindranath tagore BiographyRabindranath tagore Biography
Rabindranath tagore BiographyRaj Sikarwar
 
Hidden Surface Removal using Z-buffer
Hidden Surface Removal using Z-bufferHidden Surface Removal using Z-buffer
Hidden Surface Removal using Z-bufferRaj Sikarwar
 
Authentication in Different Scenarios
Authentication in Different ScenariosAuthentication in Different Scenarios
Authentication in Different ScenariosRaj Sikarwar
 
Authentication in Different Scenarios
Authentication in Different ScenariosAuthentication in Different Scenarios
Authentication in Different ScenariosRaj Sikarwar
 
AODV protocol and Black Hole attack
AODV protocol and Black Hole attackAODV protocol and Black Hole attack
AODV protocol and Black Hole attackRaj Sikarwar
 
Least squares support Vector Machine Classifier
Least squares support Vector Machine ClassifierLeast squares support Vector Machine Classifier
Least squares support Vector Machine ClassifierRaj Sikarwar
 

More from Raj Sikarwar (7)

Rabindranath tagore Biography
Rabindranath tagore BiographyRabindranath tagore Biography
Rabindranath tagore Biography
 
Hidden Surface Removal using Z-buffer
Hidden Surface Removal using Z-bufferHidden Surface Removal using Z-buffer
Hidden Surface Removal using Z-buffer
 
Authentication in Different Scenarios
Authentication in Different ScenariosAuthentication in Different Scenarios
Authentication in Different Scenarios
 
Authentication in Different Scenarios
Authentication in Different ScenariosAuthentication in Different Scenarios
Authentication in Different Scenarios
 
AODV protocol
AODV protocolAODV protocol
AODV protocol
 
AODV protocol and Black Hole attack
AODV protocol and Black Hole attackAODV protocol and Black Hole attack
AODV protocol and Black Hole attack
 
Least squares support Vector Machine Classifier
Least squares support Vector Machine ClassifierLeast squares support Vector Machine Classifier
Least squares support Vector Machine Classifier
 

Recently uploaded

CALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual serviceanilsa9823
 
CALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual serviceanilsa9823
 
9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girls
9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girls9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girls
9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girlsPooja Nehwal
 
call girls in candolim beach 9870370636] NORTH GOA ..
call girls in candolim beach 9870370636] NORTH GOA ..call girls in candolim beach 9870370636] NORTH GOA ..
call girls in candolim beach 9870370636] NORTH GOA ..nishakur201
 
2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)Delhi Call girls
 
2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)Delhi Call girls
 
Top Rated Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...
Top Rated  Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...Top Rated  Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...
Top Rated Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...Call Girls in Nagpur High Profile
 
8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,
8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,
8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,dollysharma2066
 
LC_YouSaidYes_NewBelieverBookletDone.pdf
LC_YouSaidYes_NewBelieverBookletDone.pdfLC_YouSaidYes_NewBelieverBookletDone.pdf
LC_YouSaidYes_NewBelieverBookletDone.pdfpastor83
 
2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)Delhi Call girls
 
CALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female service
CALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female serviceCALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female service
CALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female serviceanilsa9823
 
Lilac Illustrated Social Psychology Presentation.pptx
Lilac Illustrated Social Psychology Presentation.pptxLilac Illustrated Social Psychology Presentation.pptx
Lilac Illustrated Social Psychology Presentation.pptxABMWeaklings
 
Pokemon Go... Unraveling the Conspiracy Theory
Pokemon Go... Unraveling the Conspiracy TheoryPokemon Go... Unraveling the Conspiracy Theory
Pokemon Go... Unraveling the Conspiracy Theorydrae5
 
CALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual serviceanilsa9823
 
Call Girls Anjuna beach Mariott Resort ₰8588052666
Call Girls Anjuna beach Mariott Resort ₰8588052666Call Girls Anjuna beach Mariott Resort ₰8588052666
Call Girls Anjuna beach Mariott Resort ₰8588052666nishakur201
 
$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...
$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...
$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...PsychicRuben LoveSpells
 
Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...
Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...
Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...anilsa9823
 
CALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual serviceanilsa9823
 

Recently uploaded (20)

(Aarini) Russian Call Girls Surat Call Now 8250077686 Surat Escorts 24x7
(Aarini) Russian Call Girls Surat Call Now 8250077686 Surat Escorts 24x7(Aarini) Russian Call Girls Surat Call Now 8250077686 Surat Escorts 24x7
(Aarini) Russian Call Girls Surat Call Now 8250077686 Surat Escorts 24x7
 
CALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Jankipuram Lucknow best sexual service
 
CALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Rajajipuram Lucknow best sexual service
 
9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girls
9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girls9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girls
9892124323, Call Girls in mumbai, Vashi Call Girls , Kurla Call girls
 
call girls in candolim beach 9870370636] NORTH GOA ..
call girls in candolim beach 9870370636] NORTH GOA ..call girls in candolim beach 9870370636] NORTH GOA ..
call girls in candolim beach 9870370636] NORTH GOA ..
 
2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Dashrath Puri (Delhi)
 
2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Palam (Delhi)
 
Top Rated Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...
Top Rated  Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...Top Rated  Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...
Top Rated Pune Call Girls Tingre Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Se...
 
8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,
8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,
8377087607 Full Enjoy @24/7-CLEAN-Call Girls In Chhatarpur,
 
LC_YouSaidYes_NewBelieverBookletDone.pdf
LC_YouSaidYes_NewBelieverBookletDone.pdfLC_YouSaidYes_NewBelieverBookletDone.pdf
LC_YouSaidYes_NewBelieverBookletDone.pdf
 
2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)
2k Shots ≽ 9205541914 ≼ Call Girls In Mukherjee Nagar (Delhi)
 
CALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female service
CALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female serviceCALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female service
CALL ON ➥8923113531 🔝Call Girls Adil Nagar Lucknow best Female service
 
Lilac Illustrated Social Psychology Presentation.pptx
Lilac Illustrated Social Psychology Presentation.pptxLilac Illustrated Social Psychology Presentation.pptx
Lilac Illustrated Social Psychology Presentation.pptx
 
Pokemon Go... Unraveling the Conspiracy Theory
Pokemon Go... Unraveling the Conspiracy TheoryPokemon Go... Unraveling the Conspiracy Theory
Pokemon Go... Unraveling the Conspiracy Theory
 
CALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Mahanagar Lucknow best sexual service
 
Call Girls Anjuna beach Mariott Resort ₰8588052666
Call Girls Anjuna beach Mariott Resort ₰8588052666Call Girls Anjuna beach Mariott Resort ₰8588052666
Call Girls Anjuna beach Mariott Resort ₰8588052666
 
$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...
$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...
$ Love Spells^ 💎 (310) 882-6330 in West Virginia, WV | Psychic Reading Best B...
 
Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...
Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...
Lucknow 💋 High Class Call Girls Lucknow 10k @ I'm VIP Independent Escorts Gir...
 
(Anamika) VIP Call Girls Navi Mumbai Call Now 8250077686 Navi Mumbai Escorts ...
(Anamika) VIP Call Girls Navi Mumbai Call Now 8250077686 Navi Mumbai Escorts ...(Anamika) VIP Call Girls Navi Mumbai Call Now 8250077686 Navi Mumbai Escorts ...
(Anamika) VIP Call Girls Navi Mumbai Call Now 8250077686 Navi Mumbai Escorts ...
 
CALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual serviceCALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual service
CALL ON ➥8923113531 🔝Call Girls Aliganj Lucknow best sexual service
 

Overlapping community Detection Using Bayesian NMF

  • 1. Overlapping Community detection using Bayesian Non Negative Matrix Factorization Rajkumar Singh Rishi Barua Guide: Dr. Ashish Anand Dept. of Computer Science Indian Institute of Technology, Guwahati April 18, 2013 R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 1 / 15
  • 2. Network Paradigm R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
  • 3. Network Paradigm V ∈ RN×N is the adjacency matrix R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
  • 4. Network Paradigm V ∈ RN×N is the adjacency matrix vij can be boolean or denote connection weight R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
  • 5. Network Paradigm V ∈ RN×N is the adjacency matrix vij can be boolean or denote connection weight ki = j vij degree of node i R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
  • 6. Network Paradigm V ∈ RN×N is the adjacency matrix vij can be boolean or denote connection weight ki = j vij degree of node i data in relational form R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 2 / 15
  • 7. Commnity Detection R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 8. Commnity Detection Q. What is Community ? R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 9. Commnity Detection Q. What is Community ? As such no Specific Definition. It is Context Dependent. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 10. Commnity Detection Q. What is Community ? As such no Specific Definition. It is Context Dependent. Defn. : Here we define communities as Modules which are subgraphs with more links connecting the nodes inside than outside them. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 11. Commnity Detection Q. What is Community ? As such no Specific Definition. It is Context Dependent. Defn. : Here we define communities as Modules which are subgraphs with more links connecting the nodes inside than outside them. A given real world network is assumed to be clustered into a number of latent classes of nodes. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 12. Commnity Detection Q. What is Community ? As such no Specific Definition. It is Context Dependent. Defn. : Here we define communities as Modules which are subgraphs with more links connecting the nodes inside than outside them. A given real world network is assumed to be clustered into a number of latent classes of nodes. These nodes form regions of increased connectivity in the network. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 13. Commnity Detection Q. What is Community ? As such no Specific Definition. It is Context Dependent. Defn. : Here we define communities as Modules which are subgraphs with more links connecting the nodes inside than outside them. A given real world network is assumed to be clustered into a number of latent classes of nodes. These nodes form regions of increased connectivity in the network. These communities usually reflect functional modules that affect the overall behavior of the system. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 14. Commnity Detection Q. What is Community ? As such no Specific Definition. It is Context Dependent. Defn. : Here we define communities as Modules which are subgraphs with more links connecting the nodes inside than outside them. A given real world network is assumed to be clustered into a number of latent classes of nodes. These nodes form regions of increased connectivity in the network. These communities usually reflect functional modules that affect the overall behavior of the system. Examples: Friend cliques in social networks, Similar proteins in a protein interaction network, Research groups in a scientific collaboration network R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 3 / 15
  • 15. Community Detection R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 4 / 15
  • 16. Non-negative Matrix Factorzation Decompose data matrix V to a product of two other matrices W , H under non-negativity constraints. V ∼ ˆV = WH V , ˆV ∈ RF×N and W ∈ RF×K , H ∈ RK×N so that, FK + KN < FN Non-negativity constraints avoid the problem of an ill-posed solution They also reflect the idea of parts-based representation. V can be expressed as additive combination of certain basis structures defined by wk, given an encoding hk ˆV = k w:khk: (1) R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 5 / 15
  • 17. Poisson Model Given an Adjacency matrix V ∈ RN×N Assume, pairwise interactions vij is generated by a Poisson distribution with rate ˆvij . Hence ˆV ∼ V Expectation network ˆV is composed of W and H, so that ˆV = WH Inner rank K : Unknown number of communities wik, hkj ∈ 1, . . . , K: The contribution of each latent community to ˆvij = k wikhkj R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 6 / 15
  • 18. Cost Function R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
  • 19. Cost Function posterior: R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
  • 20. Cost Function posterior: p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2) R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
  • 21. Cost Function posterior: p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2) by taking the − log of: p(V , W , H)p(W |β)p(H|β)p(β), we define cost as U. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
  • 22. Cost Function posterior: p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2) by taking the − log of: p(V , W , H)p(W |β)p(H|β)p(β), we define cost as U. U = − log p(V |W , H) − log p(W |β) − log p(W |β) − log p(β) (3) R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
  • 23. Cost Function posterior: p(V |W , H, β) = p(V |W , H)p(W |β)p(H|β)p(β) (2) by taking the − log of: p(V , W , H)p(W |β)p(H|β)p(β), we define cost as U. U = − log p(V |W , H) − log p(W |β) − log p(W |β) − log p(β) (3) U is our minimization objective under non-negativity constraints. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 7 / 15
  • 24. Parameter inference U = i j vij log vij ˆvij + ˆvij + 1 2 k i βkw2 ik + j βkh2 kj − 2N log βk + k βkbk − (ak − 1) log βk + k R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 8 / 15
  • 25. Parameter inference U = i j vij log vij ˆvij + ˆvij + 1 2 k i βkw2 ik + j βkh2 kj − 2N log βk + k βkbk − (ak − 1) log βk + k We find w∗, h∗,β∗ that minimize U using the gradient descent algorithm. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 8 / 15
  • 26. Parameter inference U = i j vij log vij ˆvij + ˆvij + 1 2 k i βkw2 ik + j βkh2 kj − 2N log βk + k βkbk − (ak − 1) log βk + k We find w∗, h∗,β∗ that minimize U using the gradient descent algorithm. The effective number of communities K is the number of non-zero columns of W∗ and H∗ R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 8 / 15
  • 27. Results W∗, H∗ describe a bipartite network of node allocations to communities. If our original adjacency matrix V is symmetric, then W∗ = HT ∗ . Each wik or hki denotes the participation strength of node i to community k. The ith row of W or column of H describes a soft-membership distribution of node i across communities. Varying node participation scores allow us to describe overlaps between communities in a disciplined manner. R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 9 / 15
  • 28. Example: 1 R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 10 / 15
  • 29. Example: 1 R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 11 / 15
  • 30. Example: 1 R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 12 / 15
  • 31. Example: 2 We now take a 1000X1000 Adjacency matrix V , which is very sparse. The graph is a non-community graph, i.e. having no community structure. Several of the nodes are isolated. Solutions from EO, Louvian (as described in the paper) offer higher modularity than NMF. NMF clearly shows that the modularity is low and the nodes have no preference to lie in any particular community, or are non-communal. NMF does not suffer from resolution limit of modularity (which groups smaller communities), as can be clearly seen. The Output shows 1 community, with 428 isolated nodes i.e. Non-communal R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 13 / 15
  • 32. Example: 2 R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 14 / 15
  • 33. References 1 Overlapping community detection using Bayesian non-negative matrix factorization - I. Psorakis, S. Roberts and M.Ebden 2 Signal Processing with Adaptive Sparse Structured Representations - V.Tan and C.Fevotte 3 D.D. Lee and H.S. Seung, Nature ,401 R. Singh, R. Barua (IITG) Overlapping Community detection using Bayesian Non Negative Matrix FactorizationApril 18, 2013 15 / 15