Unsupervised

Learning
(a.k.a Clustering)

Marcello Pelillo
University of Venice, Italy


Artificial Intelligence
a.y. 2018/19
 The “classical” clustering problem
Given:
ü  a set of n “objects” = an edge-weighted graph G
ü  an n × n matrix A of pairwise similarities

Goal: Partition the vertices of the G into maximally homogeneous groups (i.e.,
clusters).

Usual assumption: symmetric and pairwise similarities (G is an undirected graph)

 Applications
Clustering problems abound in many areas of computer science and engineering.

A short list of applications domains:

Image processing and computer vision

Computational biology and bioinformatics
Information retrieval
Document analysis
Medical image analysis
Data mining
Signal processing
…

For a review see, e.g., A. K. Jain, "Data clustering: 50 years beyond K-means,”
Pattern Recognition Letters 31(8):651-666, 2010.
 Basic ideas of grouping in humans:
The Gestalt school

Wertheimer Gestalt properties

elements in a collection of elements
can have properties that result from
Koehler relationships
•  Gestaltqualitat

Koffka

A series of factors affect whether

elements should be grouped
together
•  Gestalt factors
Clustering
Segmentation as clustering

Source: K. Grauman
 Segmentation as clustering

•  Cluster together (pixels, tokens, •  Point-Cluster distance

etc.) that belong together –  single-link clustering
•  Agglomerative clustering –  complete-link clustering
–  attach closest to cluster it is –  group-average clustering
closest to •  Dendrograms
–  repeat –  yield a picture of output as
•  Divisive clustering clustering process continues
–  split cluster along best
boundary
–  repeat
 K-Means
An iterative clustering algorithm

– Initialize:
Pick K random points as cluster centers

– Alternate:
1. Assign data points to closest cluster center
2. Change the cluster center to the average of its assigned points

– Stop when no points’ assignments change

Note: Ensure that every cluster has at least one data point. Possible techniques for doing this include
supplying empty clusters with a point chosen at random from points far from their cluster centers.
K-means clustering: Example

Initialization:
Pick K random points as
cluster centers

Shown here for K=2

Adapted from D. Sontag

K-means clustering: Example

Iterative Step 1:
Assign data points to
closest cluster center

Adapted from D. Sontag

K-means clustering: Example

Iterative Step 2:
Change the cluster center to
the average of the assigned
points

Adapted from D. Sontag

K-means clustering: Example

Repeat until convergence

Adapted from D. Sontag

K-means clustering: Example

Final output

Adapted from D. Sontag

 Image Clusters on intensity Clusters on color

K-means clustering using intensity alone and color alone

 Properties of K-means
Guaranteed to converge in a finite number of steps.

Minimizes an objective function (compactness of clusters):

⎧ 2⎫
∑ ⎨ ∑ x j − µi ⎬
i∈clusters ⎩ j∈elements of i'th cluster ⎭

where µi is the center of cluster i.

Running time per iteration:

•  Assign data points to closest cluster center: O(Kn) time
•  Change the cluster center to the average of its points: O(n) time
 Properties of K-means

•  Pros
–  Very simple method
–  Efficient

•  Cons
–  Converges to a local minimum
of the error function
–  Need to pick K
–  Sensitive to initialization
–  Sensitive to outliers
–  Only finds “spherical” clusters
 Images as graphs

j
wij
i

•  Node for every pixel

•  Edge between every pair of pixels (or every pair of
“sufficiently close” pixels)
•  Each edge is weighted by the affinity or similarity of the
two nodes
Source: S. Seitz
 Graph-theoretic (pairwise) clustering

•  Represent tokens using a weighted graph.

–  affinity matrix
•  Cut up this graph to get subgraphs with strong interior
links
Graphs and matrices

Source: D. Sontag
 Measuring affinity

•  Suppose we represent each pixel by a feature vector

x, and define a distance function appropriate for this
feature representation

•  Then we can convert the distance between two

feature vectors into an affinity with the help of a
Gaussian kernel:

⎛ 1 2⎞
exp⎜ − 2 dist (xi , x j ) ⎟
⎝ 2σ ⎠
 Scale affects affinity
•  Small σ: group only nearby points
•  Large σ: group far-away points
 Eigenvector-based clustering
Let us represent a cluster using a vector x whose k-th entry captures the
participation of node k in that cluster. If a node does not participate in a
cluster, the corresponding entry is zero.

We also impose the restriction that xTx = 1

We want to maximize:

which is a measure for the cluster’s cohesiveness.

This is an eigenvalue problem!

Choose the eigenvector of A with largest eigenvalue
Example eigenvector

points

eigenvector

matrix
 More than two segments

•  Two options
–  Recursively split each side to get a tree, continuing till the
eigenvalues are too small
–  Use the other eigenvectors
 Clustering by eigenvectors: Algorithm

1. Construct (or take as input) the affinity matrix A

2. Compute the eigenvalues and eigenvectors of A
3. Repeat
4. Take the eigenvector corresponding to the largest unprocessed eigenvalue
5. Zero all components corresponding to elements that have already been clustered
6. Threshold the remaining components to determine which elements belong to
this cluster
7. If all elements have been accounted for, there are sufficient clusters
8. Until there are sufficient clusters
 Clustering as graph partitioning
Let G=(V, E, w) a weighted graph.

Given a “cut” (A, B), with B =V \ A, define:

cut(A, B) = ∑ ∑ w(i, j)
i∈A j∈B

A
B

Minimum Cut Problem

Among all possible cuts (A, B),

find the one which minimizes cut(A, B)
 MinCut clustering
Good news
Solvable in polynomial time

Bad news
Favors highly unbalanced clusters (often with isolated vertices)
 Graph terminology

Degree of nodes

Volume of a set

Adapted from D. Sontag

 Normalized Cut

⎛ 1 1 ⎞
Ncut(A, B) = cut(A, B) ⎜ + ⎟
⎝ vol(A) vol(B) ⎠

Adapted from D. Sontag

 Graph Laplacian (unnormalized)
Defined as

L=D–W

Example:

Assume the weights of edges are 1

 Key fact
For all vectors f in Rn, we have:

Indeed:
 Properties
•  L is symmetric (by assumption) and positive semi-definite:
f’L f ≥ 0
for all vectors f (by “key fact”)
•  Smallest eigenvalue of L is 0; corresponding eigenvector is 1
•  Thus eigenvalues are: 0 = λ1 ≤ λ2 ≤ ... ≤ λn

First relation between spectrum and clusters:

•  Multiplicity of eigenvalue λ1 = 0 is the number of connected
components of the graph
•  eigenspace is spanned by the characteristic functions of these
components (so all eigenvectors are piecewise constant)
 Normalized graph Laplacians
•  Row sum (random walk) normalization:

Lrw = D−1 L
= I – D−1 W

•  Symmetric normalization:

Lsym = D−1/2 L D−1/2

= I – D−1 W D−1/2

Spectral properties of both matrices similar to the ones of L.

 Solving Ncut
Any cut (A, B) can be represented by a binary indicator vector x:

⎧ +1 if i ∈ A
xi = ⎨
⎩ −1 if i ∈ B
Rayleigh quotient
It can be shown that:

y'(D − W )y
min Ncut(x) = min
x y y' Dy

subject to the constraint that y’D1 = ∑i yi di = 0 (with yi∈{1, -b}).

This is NP-hard!
 Ncut as an eigensystem
If we relax the constraint that y be a discrete-valued vector and allow it to
take on real values, the problem

y'(D − W )y
min
y y' Dy
is equivalent to:

min y'(D − W )y subject to y' Dy = 1

y

This amounts to solving a generalized eigenvalue problem:

Laplacian (D − W )y = λ Dy

Note: Equivalent to a standard eigenvalue problem using the normalized

Laplacian: Lrw = D−1 L = I – D−1 W.
 2-way Ncut
1.  Compute the affinity matrix W, compute the degree matrix D

2.  Solve the generalized eigenvalue problem (D – W)y = λDy

3.  Use the eigenvector associated to the second smallest eigenvalue to

bipartition the graph into two parts.

Why the second smallest eigenvalue?

Remember, the smallest eigenvalue of Laplacians is always 0
(corresponds to the trivial partition A = V, B = {})
The effect of relaxation

How to choose the splitting point?

•  Pick a constant value (0 or 0.5)

•  Pick the median value as splitting point
•  Look for the splitting point that has
minimum Ncut value:
1.  Choose n possible splitting points
2.  Compute Ncut value
3.  Pick minimum
 Random walk intepretation
Construct a Markov chain where each data point is a state, connected to
all other states with some probability.

With our affinity W and degree D, the stochastic matrix is:

P = D−1 W

which is the row-normalized version of W, so each entry P(i, j) is a

probability of “walking” to state j from state i.

Adapted from Y. Weiss

 Random walk intepretation
Problem: Finding a cut (A, B) in a graph G such that a random walk does
not have many opportunities to jump between the two clusters.

This is equivalent to the Ncut problem due to the following relation:

Ncut(A, B) = P(A | B) + P(B | A)

(Meila and Shi, 2001)

 Ncut: More than 2 clusters
Approach #1: Recursive two-way cuts

1.  Given a weighted graph G = (V, E, w), summarize the information into
matrices W and D

2.  Solve (D − W)y = λDy for eigenvectors with the smallest eigenvalues

3.  Use the eigenvector with the second smallest eigenvalue to bipartition the
graph by finding the splitting point such that Ncut is minimized

4.  Decide if the current partition should be subdivided by checking the stability of
the cut, and make sure Ncut is below the prespecified value

5.  Recursively repartition the segmented parts if necessary

Note. The approach is computationally wasteful; only the second eigenvector is used, whereas
the next few small eigenvectors also contain useful partitioning information.
 Ncut: More than 2 clusters
Approach #2: Using first k eigenvectors
Spectral clustering

Ng, Jordan and Weiss (2002)

 K-means vs Spectral clustering
Applying k-means to Laplacian eigenvectors allows us to find cluster with
non-convex boundaries.

Adapted from A. Singh

 K-means vs Spectral clustering
Applying k-means to Laplacian eigenvectors allows us to find cluster with
non-convex boundaries.

Adapted from A. Singh

 K-means vs Spectral clustering
Applying k-means to Laplacian eigenvectors allows us to find cluster with
non-convex boundaries.

Adapted from A. Singh

Examples
Examples (choice of k)
 Choosing k
The eigengap heuristic: Choose k such that all eigenvalues λ1,…, λk are
very small, but λk+1 is relatively large

Four 1D Gaussian clusters with increasing variance and corresponding eigevalues of Lrw (von Luxburg, 2007).
 References
•  J. Shi and J. Malik, Normalized cuts and image segmentation. IEEE
Transactions on Pattern Analysis and Machine Intelligence 22(8): 888-905
(2000).

•  M. Meila and J. Shi. A random walks view of spectral segmentation. AISTATS

(2001).

•  A. Ng, M. Jordan, and Y. Weiss. On spectral clustering: Analysis and

analgorithm. NIPS 14 (2002).

•  U. von Luxburg, A tutorial on spectral clustering. Statistics and Computing

17(4) 395-416 (2007).

•  A. K. Jain, Data clustering: 50 years beyond K-means. Pattern Recognition

Letters 31(8):651-666 (2010).
Dominant Sets
The Need for Non-exhaustive Clusterings
Separating Structure from Clutter
Separating Structure from Clutter
NCut

K-means Our approach

 One-class Clustering

“[…] in certain real-world problems, natural groupings are found

among only on a small subset of the data, while the rest of the data
shows little or no clustering tendencies.
In such situations it is often more important to cluster a small
subset of the data very well, rather than optimizing a clustering
criterion over all the data points, particularly in application
scenarios where a large amount of noisy data is encountered.”

G. Gupta and J. Ghosh. Bregman bubble clustering: A robust framework

for mining dense cluster. ACM Trans. Knowl. Discov. Data (2008).
 When Groups Overlap

Does O belong to AD or to BC (or to none)?

 The Need for Overlapping Clusters
Partitional approaches impose that each element cannot belong to more than one
cluster. There are a variety of important applications, however, where this
requirement is too restrictive.

Examples:
ü  clustering micro-array gene expression data
ü  clustering documents into topic categories
ü  perceptual grouping
ü  segmentation of images with transparent surfaces

References:
ü  N. Jardine and R. Sibson. The construction of hierarchic and non-hierarchic
classifications. Computer Journal, 11:177–184, 1968
ü  A. Banerjee, C. Krumpelman, S. Basu, R. J. Mooney, and J. Ghosh. Model-
based overlapping clustering. KDD 2005.
ü  K. A. Heller and Z. Ghahramani. A nonparametric Bayesian approach to
modeling overlapping clusters. AISTATS 2007.
 The Symmetry Assumption
«Similarity has been viewed by both philosophers and psychologists
as a prime example of a symmetric relation. Indeed, the assumption
of symmetry underlies essentially all theoretical treatments of
similarity.

Contrary to this tradition, the present paper provides empirical

evidence for asymmetric similarities and argues that similarity should
not be treated as a symmetric relation.»

Amos Tversky
“Features of similarities,” Psychol. Rev. (1977)

Examples of asymmetric (dis)similarities

ü  Kullback-Leibler divergence
ü  Directed Hausdorff distance
ü  Tversky’s contrast model
 What is a Cluster?
No universally accepted (formal) definition of a “cluster” but, informally, a
cluster should satisfy two criteria:

Internal criterion
all “objects” inside a cluster should be highly similar to each other

External criterion
all “objects” outside a cluster should be highly dissimilar to the ones inside
 The Notion of “Gestalt”

«In most visual fields the contents of particular areas “belong together” as
circumscribed units from which their surrounding are excluded.»

W. Köhler, Gestalt Psychology (1947)

«In gestalt theory the word “Gestalt” means any segregated whole.»

W. Köhler (1929)
 Data Clustering:
Old vs. New
By answering the question “what is a cluster?” we get a novel way of
looking at the clustering problem.

Clustering_old(V,A,k)
V1,V2,...,Vk <- My_favorite_partitioning_algorithm(V,A,k)
return V1,V2,...,Vk

−−−−−−

Clustering_new(V,A)
V1,V2,...,Vk <- Enumerate_all_clusters(V,A)
return V1,V2,...,Vk

Enumerate_all_clusters(V,A)
repeat
Extract_a_cluster(V,A)
until all clusters have been found
return the clusters found
 A Special Case:
Binary Symmetric Affinities

Suppose the similarity matrix is a binary (0/1) matrix.

Given an unweighted undirected graph G=(V,E):

A clique is a subset of mutually adjacent vertices

A maximal clique is a clique that is not contained in a larger one

In the 0/1 case, a meaningful notion of a cluster is that of a maximal clique.

NCut

New approach
 Advantages of the New Approach

ü  No need to know the number of clusters in advance (since we extract

them sequentially)

ü  Leaves clutter elements unassigned (useful, e.g., in figure/ground

separation or one-class clustering problems)

ü  Allows extracting overlapping clusters

Need a partition?

Partition_into_clusters(V,A)
repeat
Extract_a_cluster
remove it from V
until all vertices have been clustered
 What is Game Theory?
“The central problem of game theory was posed by von
Neumann as early as 1926 in Göttingen. It is the following:
If n players, P1,…, Pn, play a given game Γ, how must the ith
player, Pi, play to achieve the most favorable result for himself?”

Harold W. Kuhn
Lectures on the Theory of Games (1953)

A few cornerstones in game theory

1921−1928: Emile Borel and John von Neumann give the first modern formulation of a
mixed strategy along with the idea of finding minimax solutions of normal-form games.

1944, 1947: John von Neumann and Oskar Morgenstern publish Theory of Games and
Economic Behavior.

1950−1953: In four papers John Nash made seminal contributions to both non-cooperative
game theory and to bargaining theory.

1972−1982: John Maynard Smith applies game theory to biological problems thereby
founding “evolutionary game theory.”

late 1990’s −: Development of algorithmic game theory…

 “Solving” a Game

Player 2

Left Middle Right

Top 3,1 2,3 10 , 2

High 4,5 3,0 6,4

Player 1
Low 2,2 5,4 12 , 3

Bottom 5,6 4,5 9,7

 Basics of (Two-Player, Symmetric)
Game Theory
Assume:
–  a (symmetric) game between two players
–  complete knowledge
–  a pre-existing set of pure strategies (actions) O={o1,…,on} available
to the players.

Each player receives a payoff depending on the strategies selected by him

and by the adversary. Players’ goal is to maximize their own returns.

A mixed strategy is a probability distribution x=(x1,…,xn)T over the strategies.

⎧ n ⎫
Δ = ⎨ x ∈ R : ∀i = 1…n : x i ≥ 0, and ∑ x i = 1⎬
n

⎩ i=1 ⎭
 Nash Equilibria

ü  Let A be an arbitrary payoff matrix: aij is the payoff obtained by playing i

while the opponent plays j.

ü  The average payoff obtained by playing mixed strategy y while the

opponent plays x, is:

y ʹAx = ∑ ∑ aij y i x j
i j

ü  A mixed strategy x is a (symmetric) Nash equilibrium if

€ x'Ax ≥ y ʹAx
for all strategies y. (Best reply to itself.)

Theorem (Nash,
€ 1951). Every finite normal-form game admits a mixed-
strategy Nash equilibrium.
 Evolution and the Theory of Games

“We repeat most emphatically that our theory is thoroughly static.

A dynamic theory would unquestionably be more complete and
therefore preferable.
But there is ample evidence from other branches of science that it
is futile to try to build one as long as the static side is not
thoroughly understood.”

John von Neumann and Oskar Morgenstern

Theory of Games and Economic Behavior (1944)

“Paradoxically, it has turned out that game theory is more readily

applied to biology than to the field of economic behaviour for
which it was originally designed.”

John Maynard Smith

Evolution and the Theory of Games (1982)
 Evolutionary Games and ESS’s

Assumptions:
ü  A large population of individuals belonging to the same species which
compete for a particular limited resource
ü  This kind of conflict is modeled as a symmetric two-player game, the
players being pairs of randomly selected population members
ü  Players do not behave “rationally” but act according to a pre-
programmed behavioral pattern (pure strategy)
ü  Reproduction is assumed to be asexual
ü  Utility is measured in terms of Darwinian fitness, or reproductive
success

A Nash equilibrium x is an Evolutionary Stable Strategy (ESS) if, for all

strategies y:
 ESS’s as Clusters

We claim that ESS’s abstract well the main characteristics of a cluster:

ü  Internal coherency: High mutual support of all elements within the

group.

ü  External incoherency: Low support from elements of the group to

elements outside the group.
 Basic Definitions

Let S ⊆ V be a non-empty subset of vertices, and i∈S.

The (average) weighted degree of i w.r.t. S is defined as:

1
awdeg S (i) = ∑ aij
| S | j ∈S

Moreover,
€ if j ∉ S, we define:
i j
φ S (i, j) = aij − awdeg S (i)
S

Intuitively,
€ φS(i,j) measures the similarity between vertices j and i, with
respect to the (average) similarity between vertex i and its neighbors in S.
 Assigning Weights to Vertices

Let S ⊆ V be a non-empty subset of vertices, and i∈S.

The weight of i w.r.t. S is defined as:

⎧1 if S = 1
⎪
w S (i) = ⎨ ∑ φ ( j,i)w S−{ i} ( j) otherwise
S−{ i}
⎪⎩ j ∈S−{ i} S-{i}

€ Further, the total weight of S is defined as:

j

i
W (S) = ∑ w S (i)
i∈S
S
 Interpretation

Intuitively, wS(i) gives us a measure of the overall (relative) similarity between

vertex i and the vertices of S-{i} with respect to the overall similarity among the
vertices in S-{i}.

w{1,2,3,4}(1) < 0 w{1,2,3,4}(1) > 0

 Dominant Sets

Definition (Pavan and Pelillo, 2003, 2007). A non-empty subset of vertices S ⊆

V such that W(T) > 0 for any non-empty T ⊆ S, is said to be a dominant set if:

1.  wS(i) > 0, for all i ∈ S (internal homogeneity)

2.  wS∪{i}(i) < 0, for all i ∉ S (external homogeneity)

Dominant sets ≡ clusters

The set {1,2,3} is dominant.

 The Clustering Game
Consider the following “clustering game.”

ü  Assume a preexisting set of objects O and a (possibly asymmetric) matrix

of affinities A between the elements of O.
ü  Two players play by simultaneously selecting an element of O.
ü  After both have shown their choice, each player receives a payoff
proportional to the affinity that the chosen element has wrt the element
chosen by the opponent.

Clearly, it is in each player’s interest to pick an element that is strongly

supported by the elements that the adversary is likely to choose.

Hence, in the (pairwise) clustering game:

ü  There are 2 players (because we have pairwise affinities)

ü  The objects to be clustered are the pure strategies
ü  The (null-diagonal) affinity matrix coincides with the similarity matrix
 Dominant Sets are ESS’s

Theorem (Torsello, Rota Bulò and Pelillo, 2006). Evolutionary stable strategies
of the clustering game with affinity matrix A are in a one-to-one
correspondence with dominant sets.

Note. Generalization of well-known Motzkin-Straus theorem from graph theory

(1965).

Dominant-set clustering

ü  To get a single dominant-set cluster use, e.g., replicator dynamics (but see
Rota Bulò, Pelillo and Bomze, CVIU 2011, for faster dynamics)

ü  To get a partition use a simple peel-off strategy: iteratively find a dominant

set and remove it from the graph, until all vertices have been clustered

ü  To get overlapping clusters, enumerate dominant sets (see Bomze, 1992;

Torsello, Rota Bulò and Pelillo, 2008)
 Special Case:
Symmetric Affinities

Given a symmetric real-valued matrix A (with null diagonal), consider the

following Standard Quadratic Programming problem (StQP):

maximize ƒ(x) = xTAx

subject to x∈∆

Note. The function ƒ(x) provides a measure of cohesiveness of a cluster (see

Pavan and Pelillo, 2003, 2007; Sarkar and Boyer, 1998; Perona and Freeman,
1998).

ESS’s are in one-to-one correspondence

to (strict) local solutions of StQP

Note. In the 0/1 (symmetric) case, ESS’s are in one-to-one correspondence to

(strictly) maximal cliques (Motzkin-Straus theorem).
 Replicator Dynamics
Let xi(t) the population share playing pure strategy i at time t. The state of the
population at time t is: x(t)= (x1(t),…,xn(t))∈∆.

Replicator dynamics (Taylor and Jonker, 1978) are motivated by Darwin’s

principle of natural selection:
x˙ i
∝ payoff of pure strategy i − average population payoff
xi

which yields:

€
x˙ i = x i [(Ax) i − x T Ax ]

Theorem (Nachbar, 1990; Taylor and Jonker, 1978). A point x∈∆ is a Nash
equilibrium if and only if x is the limit point of a replicator dynamics
€ from the interior of ∆.
trajectory starting
Furthermore, if x∈∆ is an ESS, then it is an asymptotically stable equilibrium
point for the replicator dynamics.
 Doubly Symmetric Games
In a doubly symmetric (or partnership) game, the payoff matrix A is
symmetric (A = AT).

Fundamental Theorem of Natural Selection (Losert and Akin, 1983).

For any doubly symmetric game, the average population payoff ƒ(x) =
xTAx is strictly increasing along any non-constant trajectory of replicator
dynamics, namely, d/dtƒ(x(t)) ≥ 0 for all t ≥ 0, with equality if and only if
x(t) is a stationary point.

Characterization of ESS’s (Hofbauer and Sigmund, 1988)

For any doubly simmetric game with payoff matrix A, the following
statements are equivalent:
a)  x ∈ ∆ESS
b)  x ∈ ∆ is a strict local maximizer of ƒ(x) = xTAx over the standard
simplex ∆
c)  x ∈ ∆ is asymptotically stable in the replicator dynamics
 Discrete-time Replicator Dynamics
A well-known discretization of replicator dynamics, which assumes non-
overlapping generations, is the following (assuming a non-negative A):

A( x(t)) i
x i (t + 1) = x i (t)
x(t)T Ax(t)

which inherits most of the dynamical properties of its continuous-time

counterpart (e.g., the fundamental theorem of natural selection).
€
MATLAB implementation
 A Toy Example

 Measuring the Degree of Cluster
Membership
The components of the converged vector give us a measure of the participation of
the corresponding vertices in the cluster, while the value of the objective function
provides of the cohesiveness of the cluster.
 Application to Image Segmentation

An image is represented as an edge-weighted undirected graph, where

vertices correspond to individual pixels and edge-weights reflect the
“similarity” between pairs of vertices.

For the sake of comparison, in the experiments we used the same similarities
used in Shi and Malik’s normalized-cut paper (PAMI 2000).

To find a hard partition, the following peel-off strategy was used:

Partition_into_dominant_sets(G)
Repeat
find a dominant set
remove it from graph
until all vertices have been clustered

To find a single dominant set we used replicator dynamics (but see Rota
Bulò, Pelillo and Bomze, CVIU 2011, for faster game dynamics).
 Intensity Segmentation Results

Dominant sets Ncut

 Intensity Segmentation Results

Dominant sets Ncut

Results on the Berkeley Dataset
Dominant sets Ncut
 Color Segmentation Results

Original image Dominant sets Ncut

Results on the Berkeley Dataset
Dominant sets Ncut
Texture Segmentation Results

Dominant sets
Texture Segmentation Results

NCut
 In a nutshell…
The game-theoretic/dominant-set approach:

ü  makes no assumption on the structure of the affinity matrix, being it able to

work with asymmetric and even negative similarity functions

ü  does not require a priori knowledge on the number of clusters (since it extracts
them sequentially)

ü  leaves clutter elements unassigned (useful, e.g., in figure/ground separation or

one-class clustering problems)

ü  allows principled ways of assigning out-of-sample items (NIPS’04)

ü  allows extracting overlapping clusters (ICPR’08)

ü  generalizes naturally to hypergraph clustering problems, i.e., in the presence

of high-order affinities, in which case the clustering game is played by more
than two players (PAMI’13)

ü  extends to hierarchical clustering (ICCV’03: EMMCVPR’09)

ü  allows using multiple affinity matrices using Pareto-Nash notion (SIMBAD’15)

 References
S. Rota Bulò and M. Pelillo. Dominant-set clustering: A review. EJOR (2017).
M. Pavan and M. Pelillo. Dominant sets and pairwise clustering. PAMI 2007.
S. Rota Bulò and M. Pelillo. A game-theoretic approach to hypergraph clustering.
PAMI’13.
A. Torsello, S. Rota Bulò and M. Pelillo. Grouping with asymmetric affinities: A
game-theoretic perspective. CVPR 2006.
A. Torsello, S. Rota Bulò and M. Pelillo. Beyond partitions: Allowing overlapping
groups in pairwise clustering. ICPR 2008.
M. Pelillo. What is a cluster? Perspectives from game theory. NIPS 2009 Workshop
on “Clustering: Science or Art?” (talk available on videolectures.net).
S. Rota Bulò, M. Pelillo and I. M. Bomze. Graph-based quadratic optimization: A
fast evolutionary approach. CVIU 2011.
S. Vascon et al., Detecting conversational groups in images and sequences: A
robust game-theoretic approach. CVIU 2016.
E. Zemene and M. Pelillo, Interactive image segmentation using constrained
dominant sets. ECCV 2016.