SlideShare ist ein Scribd-Unternehmen logo
1 von 51
Downloaden Sie, um offline zu lesen
@tksakaki

#nipsreading
              1

 


    Twitter
     ◦ 
 
  Properties of Clustering Functions
  A taxonomy of k-clustering fucntions
 

 
 
  Properties of Clustering Functions
  A taxonomy of k-clustering fucntions
 

 
Clustering:
                             ↓
       Clustering

                    Ad-hoc


               	
○Clustering

○     Clustering             property
Clustering




         clustering
    A Impossibility Theorem for Clustering
     ◦  Jon Kleinberg, NIPS 2002


    Measures of Clustering Quality: A Working Set
     of Axioms for Clustering
     ◦  M.Ackerman and S.Ben-David, NIPS 2008


    Characterization of Linkage-based
     Clustering.
     ◦  M.Ackerman and S.Ben-David, COLT 2010
X:                              	
                              +
            	
d : X   × X →R ,d ( x, x ) = 0(∀x ∈ X )
                 	
 X,d )
                 (         clustering      	
 ( X,d,k )
                                           F

    clustering        	
€        C = F ( X,d,k )
    €                 ⎛   €               ⎞
         {C1,C2 ,Ck }⎜ Ci = X,1 ≤ k ≤ X ⎟
                      ⎝ i                 ⎠
general clustering function F	
Input:   F ( X,d )
                           ⎛                   ⎞
Output: 	
C = {C1,C2 ,Ck }⎜ Ci = X,1 ≤ k ≤ X ⎟
                              ⎝   i             ⎠

    €
 k-clustering function F	
   €

Input: F ( X,d,k ), (1 ≤ k ≤ X )
                           ⎛                  ⎞
Output: 	
 = {C1,C2 ,Ck }⎜ Ci = X,1 ≤ k ≤ X ⎟
         C
                           ⎝ i                ⎠

€
    €
 
  Properties of Clustering Functions
  A taxonomy of k-clustering fucntions
 

 
iso. invariance	
               	
             scale invariance	
                          	
       	
             order invariance	
             outer consistency	
   cluster                          	
             inner consistency	
   cluster                     	
       	
             k-rich	
                                   k                                	
             inner rich	
richness	




                                                                                    	
             outer rich	
                                                           	
             threshold rich	
                                                 	
             locality	
            clustering                            	
             refinement-confined	
 k              clustering
clustering       	

φ : X → X ʹ′
x, y ∈ X,d ( x, y ) = dʹ′(φ (x), φ (y))
F ( X,d,k ),F ( X ʹ′, dʹ′,k ) : isomorphic(∀k)
               x, y : same → φ (x), φ (y) : same
clustering
clustering   	


x, y ∈ X,
d ( x, y ) = cdʹ′( x, y )
→F ( X,d,k ) = F ( X ʹ′, dʹ′,k )
2                                            clustering                  	

        x1, x 2 , x 3 , x 4 ∈ X,
        d ( x1, x 2 ) < d ( x 3 , x 4 ), dʹ′( x1, x 2 ) < dʹ′( x 3 , x 4 )
        →F ( X,d,k ) = F ( X ʹ′, dʹ′,k )



€
2                          clustering       	


            (Single-linkage clustering 	




    0   1    4   9   10 12 15   19   20
clustering      cluster
             cluster

                          C’
clustering      cluster
             cluster


   C = F(X,d,k),
    Cʹ′ ⊆ C
    F(Cʹ′,d,| Cʹ′ |)= Cʹ′
cluster
      cluster                           clustering
   cluster                    cluster


d(x,y)
                    d’(x,y)
                                    d(x,y)


                                         d’(x,y)
cluster
      cluster                       clustering
   cluster                cluster

dʹ′ : (C,d ) − consistent

x, y : same → dʹ′( x, y ) ≤ d ( x, y )
x, y : different → dʹ′( x, y ) ≥ d ( x, y )
cluster                        clustering
          cluster




                    d(x,y)

                         d’(x,y)
cluster                 clustering
              cluster


dʹ′ : (C,d) − outerconsistent
x, y : same → dʹ′( x, y ) = d ( x, y )
x, y : different → dʹ′( x, y ) ≥ d ( x, y )
cluster             clustering
          cluster




     d(x,y)

                       d’(x,y)
cluster                clustering
               cluster


dʹ′ : (C,d) − innterconsistent
x, y : same → dʹ′( x, y ) ≤ d ( x, y )
x, y : different → dʹ′( x, y ) = d ( x, y )
clustering



any : X1, X 2  X k
X ʹ′ = { X1, X 2  X k }
→∃d : F ( X ʹ′,d,k ) = { X1, X 2  X k }
clustering


※                                        clustering



    any : (X1,d1 ),(X 2 ,d2 )(X k ,dk )
             ⎛ k          ⎞
    →∃d : F ⎜ X i , d,k ⎟ = { X1, X 2  X k }
         ˆ             ˆ
             ⎝ i=1        ⎠
    ˆ
    d : entends − d (i ≤ k)
                     i
clustering


※                                                 clustering



    ( X,d), X = { X1, X 2  X k }
      ˆ              ˆ           (
    →∃d : d ( a,b) = d ( a,b) a ∈ X i ,b ∈ X j ,i ≠ j          )
      ⎛ k           ⎞
    F ⎜  X i , d,k ⎟ = { X1, X 2  X k }
                 ˆ
      ⎝ i=1         ⎠
clustering




∃a < b
x, y : same →d(x, y) ≤ a,
x, y : different →d(x, y) ≥ b
F ( X,d, C ) = C
k≦k’       F(X,d,k’)   F(X,d,k)
                                refinement




    1 ≤ k ≤ kʹ′ ≤ X ,
    O( F ( X,d,k')) ≥ O( F ( X,d,k ))



€
k≦k’   F(X,d,k’)   F(X,d,k)
                       refinement
iso. invariance	
               	
             scale invariance	
                          	
       	
             order invariance	
             outer consistency	
   cluster                          	
             inner consistency	
   cluster                     	
       	
             k-rich	
                                   k                                	
             inner rich	
richness	




                                                                                    	
             outer rich	
                                                           	
             threshold rich	
                                                 	
             locality	
            clustering                            	
             refinement-confined	
 k              clustering
    Clustering                Kleinberg 	

                  clustering




                     Scale-Invariance
                         Richness
                       Consistency
    single linkage clustering
                     stop condition
    Consistency + Richness: only link if distance is less than r
     ◦                                      clustering
                        cluster

    Consistency + SI: stop when you have k connected
     components
     ◦                                  clustering      /
                   clustering

    Richness + SI: if x is the diameter of the graph, only add
     edges with weight βx
     ◦                          cluster     /
         clustering
 
  Properties of Clustering Functions
  A taxonomy of k-clustering fucntions
 

 
outer consistent	

                                       inner consistent	




                                                                                                                                                                   scale invariant 	
                                                                                     order invariant	




                                                                                                                                                threshold rich	




                                                                                                                                                                                        iso. invariant	
                                                                      refinement	




                                                                                                                                 inner rich	
                                                                                                                    out rich	
                                                                                                         k-rich	
                                                            local	
Singe Linkage	
    ○	
                  ○	
                 ○	
       ○	
             ○	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Average Linkage    ○	
                  ×	
                 ○	
       ○	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Complete Linkage	
 ○	
                  ×                   ○	
       ○	
             ○                  ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-median	
         ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-means	
          ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Min sum	
          ○	
                  ○	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Ratio cut	
        ×	
                  ○	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Nomalize cut	
     ×	
                  ×	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○
outer consistent	

                                       inner consistent	




                                                                                                                                                                   scale invariant 	
                                                                                     order invariant	




                                                                                                                                                threshold rich	




                                                                                                                                                                                        iso. invariant	
                                                                      refinement	




                                                                                                                                 inner rich	
                                                                                                                    out rich	
                                                                                                         k-rich	
                                                            local	
Singe Linkage	
    ○	
                  ○	
                 ○	
       ○	
             ○	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Average Linkage    ○	
                  ×	
                 ○	
       ○	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Complete Linkage	
 ○	
                  ×                   ○	
       ○	
             ○                  ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-median	
         ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-means	
          ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Min sum	
          ○	
                  ○	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Ratio cut	
        ×	
                  ○	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Nomalize cut	
     ×	
                  ×	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○
outer consistent	

                                       inner consistent	




                                                                                                                                                                   scale invariant 	
                                                                                     order invariant	




                                                                                                                                                threshold rich	




                                                                                                                                                                                        iso. invariant	
                                                                      refinement	




                                                                                                                                 inner rich	
                                                                                                                    out rich	
                                                                                                         k-rich	
                                                            local	
Singe Linkage	
    ○	
                  ○	
                 ○	
       ○	
             ○	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Average Linkage    ○	
                  ×	
                 ○	
       ○	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Complete Linkage	
 ○	
                  ×                   ○	
       ○	
             ○                  ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-median	
         ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-means	
          ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Min sum	
          ○	
                  ○	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Ratio cut	
        ×	
                  ○	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Nomalize cut	
     ×	
                  ×	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○
outer consistent	

                                       inner consistent	




                                                                                                                                                                   scale invariant 	
                                                                                     order invariant	




                                                                                                                                                threshold rich	




                                                                                                                                                                                        iso. invariant	
                                                                      refinement	




                                                                                                                                 inner rich	
                                                                                                                    out rich	
                                                                                                         k-rich	
                                                            local	
Singe Linkage	
    ○	
                  ○	
                 ○	
       ○	
             ○	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Average Linkage    ○	
                  ×	
                 ○	
       ○	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Complete Linkage	
 ○	
                  ×                   ○	
       ○	
             ○                  ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-median	
         ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

k-means	
          ○	
                  ×	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Min sum	
          ○	
                  ○	
                 ○	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Ratio cut	
        ×	
                  ○	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Nomalize cut	
     ×	
                  ×	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○
outer consistent	

                                        inner consistent	




                                                                                                                                                                    scale invariant 	
                                                                                      order invariant	




                                                                                                                                                 threshold rich	




                                                                                                                                                                                         iso. invariant	
                                                                       refinement	




                                                                                                                                  inner rich	
                                                                                                                     out rich	
                                                                                                          k-rich	
                                                             local	
Singe Linkage	
     ○	
                  ○	
                 ○	
       ○	
             ○	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Average Linkage     ○	
                  ×	
                 ○	
       ○	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Complete Linkage	
 clustering○	
                   ○	
  ×                                              ○	
             ○                  ○	
        ○	
          ○	
             ○	
                ○	
                  ○	
 scale invariance
k-median	
   ○	
 ×	
 ○	
                         : ×	
 ×	
                           natural                                                                        ○	
        ○	
          ○	
             ○	
                ○	
                  ○	
 isomorphism variance :natural
k-means	
    ○	
 ×	
 ○	
   ×	
 ×	
                                                                        ○	
        ○	
          ○	
             ○	
                ○	
                  ○	
 threshold richness
Min sum	
           ○	
                  ○	
                 ○	
       ×	
 / ×	
                          ○	
        ○	
          ○	
             ○	
                ○	
                  ○	
              clustering
Ratio cut	
         ×	
                  ○	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○	

Nomalize cut	
      ×	
                  ×	
                 ×	
       ×	
             ×	
                ○	
        ○	
          ○	
             ○	
                ○	
                  ○
 
  Properties of Clustering Functions
  A taxonomy of k-clustering fucntions
 

 
  Invariance properties
      Consistency properties
            (C,d) − nice var iant
                [                    ] [
            P F ( X, dʹ′, C ) = C ≥ P F ( X,d, C ) = C     ]
        Richness properties
            ∀ε > 0
€
            ∃d : P ( F ( X,d,k ) = C ) ≥1 − ε
        Locality	
                [                          ]
            P F ( X ʹ′,d / X ʹ′, Cʹ′ ) = Cʹ′

€
            =
                    [
                P Cʹ′ ⊆ C F ( X,d, j ) = CandC / X ʹ′isak − clustering    ]
                         [                                        ]
                        P ∃C1,C2 Ck s.t.Ci = X ʹ′ F ( X,d, j ) = C ≠ 0
    k-means
    k-means
Random Centroids Lloyd	

                                        	
Furthest Centroids Lloyd	



                                    (
            maximizemin1≤ j ≤i−1 d c j ,c i   )
                                                  	
 €
Clustering Algorithm	




                         outer consistent	




                                                                           scale invariant	
                                                        threshold rich	




                                                                                               iso. invariant	




                                                                                                                             outer rich	
                                                                                                                  k-rich	
                                              local	
Optimal k-means	
         ○	
 ○	
 ○	
 ○	
 ○	
 ○	
 ○	
Random Centroid Lloyd	
 ×	
 ×	
 ×	
 ○	
 ○	
 ○	
 ○	
Furthest Centroid Lloyd	
 ×	
 ×	
 ○	
 ○	
 ○	
 ○	
 ○
Clustering Algorithm	




                         outer consistent	




                                                                           scale invariant	
                                                        threshold rich	




                                                                                               iso. invariant	




                                                                                                                             outer rich	
                                                                                                                  k-rich	
                                              local	
Optimal k-means	
         ○	
 ○	
 ○	
 ○	
 ○	
 ○	
 ○	
Random Centroid Lloyd	
 ×	
 ×	
 ×	
 ○	
 ○	
 ○	
 ○	
Furthest Centroid Lloyd	
 ×	
 ×	
 ○	
 ○	
 ○	
 ○	
 ○	

threshold richness                Furthest Centroid
Lloyd       Random Centroid Lloyd
    Kleinberg            	

                       clustering   	




                 Scale-Invariance
                     Richness
                   Consistency
            	

            clustering     	




      Scale-Invariance
          Richness
     Outer-Consistency
 
  Properties of Clustering Functions
  A taxonomy of k-clustering fucntions
 

 
Clustering Function property
                                 	



     clustering axioms       scale-invariance,
isomorphism-invariance, threshold richness




    Kleinberg
    Supervised Clustering
     ◦  2008                   clustering
     ◦                    clustering           merge/
        split

    Efficient Robust Feature Selection via Joint
     L2,1-Norms Minimization
     ◦  Bio Informatics          Feature Selection
     ◦  L1,2-norm SVM           Feature

Weitere ähnliche Inhalte

Was ist angesagt?

S101-52國立新化高中(代理)
S101-52國立新化高中(代理)S101-52國立新化高中(代理)
S101-52國立新化高中(代理)
yustar1026
 
Bt0063 mathematics fot it
Bt0063 mathematics fot itBt0063 mathematics fot it
Bt0063 mathematics fot it
nimbalkarks
 
Computer Science and Information Science 3rd semester (2012-December) Questio...
Computer Science and Information Science 3rd semester (2012-December) Questio...Computer Science and Information Science 3rd semester (2012-December) Questio...
Computer Science and Information Science 3rd semester (2012-December) Questio...
B G S Institute of Technolgy
 
09 Trial Penang S1
09 Trial Penang S109 Trial Penang S1
09 Trial Penang S1
guest9442c5
 
Ecuaciones ejercicios de repaso-GRADO 10°-2013
Ecuaciones ejercicios de repaso-GRADO 10°-2013Ecuaciones ejercicios de repaso-GRADO 10°-2013
Ecuaciones ejercicios de repaso-GRADO 10°-2013
Edwin Rivera Cantor
 

Was ist angesagt? (16)

Mathematics
MathematicsMathematics
Mathematics
 
S101-52國立新化高中(代理)
S101-52國立新化高中(代理)S101-52國立新化高中(代理)
S101-52國立新化高中(代理)
 
Elliptic Curve Cryptography and Zero Knowledge Proof
Elliptic Curve Cryptography and Zero Knowledge ProofElliptic Curve Cryptography and Zero Knowledge Proof
Elliptic Curve Cryptography and Zero Knowledge Proof
 
Bt0063 mathematics fot it
Bt0063 mathematics fot itBt0063 mathematics fot it
Bt0063 mathematics fot it
 
Chapter 15
Chapter 15Chapter 15
Chapter 15
 
Classification Theory
Classification TheoryClassification Theory
Classification Theory
 
NCCU CPDA Lecture 12 Attribute Based Encryption
NCCU CPDA Lecture 12 Attribute Based EncryptionNCCU CPDA Lecture 12 Attribute Based Encryption
NCCU CPDA Lecture 12 Attribute Based Encryption
 
Antiderivatives nako sa calculus official
Antiderivatives nako sa calculus officialAntiderivatives nako sa calculus official
Antiderivatives nako sa calculus official
 
Applied Calculus Chapter 4 multiple integrals
Applied Calculus Chapter  4 multiple integralsApplied Calculus Chapter  4 multiple integrals
Applied Calculus Chapter 4 multiple integrals
 
Semantic Data Box
Semantic Data BoxSemantic Data Box
Semantic Data Box
 
Cheat Sheet
Cheat SheetCheat Sheet
Cheat Sheet
 
Decimal To Hexadecimal Conversion Tips
Decimal To Hexadecimal Conversion TipsDecimal To Hexadecimal Conversion Tips
Decimal To Hexadecimal Conversion Tips
 
Sect1 2
Sect1 2Sect1 2
Sect1 2
 
Computer Science and Information Science 3rd semester (2012-December) Questio...
Computer Science and Information Science 3rd semester (2012-December) Questio...Computer Science and Information Science 3rd semester (2012-December) Questio...
Computer Science and Information Science 3rd semester (2012-December) Questio...
 
09 Trial Penang S1
09 Trial Penang S109 Trial Penang S1
09 Trial Penang S1
 
Ecuaciones ejercicios de repaso-GRADO 10°-2013
Ecuaciones ejercicios de repaso-GRADO 10°-2013Ecuaciones ejercicios de repaso-GRADO 10°-2013
Ecuaciones ejercicios de repaso-GRADO 10°-2013
 

Ähnlich wie nips勉強会_Toward Property-Based Classification of Clustering Paradigms

WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...
WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...
WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...
grssieee
 
An introduction to quantum stochastic calculus
An introduction to quantum stochastic calculusAn introduction to quantum stochastic calculus
An introduction to quantum stochastic calculus
Springer
 
TunUp final presentation
TunUp final presentationTunUp final presentation
TunUp final presentation
Gianmario Spacagna
 
Lesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of CalculusLesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of Calculus
Matthew Leingang
 
Lesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of CalculusLesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of Calculus
Matthew Leingang
 
Information-theoretic clustering with applications
Information-theoretic clustering  with applicationsInformation-theoretic clustering  with applications
Information-theoretic clustering with applications
Frank Nielsen
 
Triangle counting handout
Triangle counting handoutTriangle counting handout
Triangle counting handout
csedays
 

Ähnlich wie nips勉強会_Toward Property-Based Classification of Clustering Paradigms (20)

Complex and Social Network Analysis in Python
Complex and Social Network Analysis in PythonComplex and Social Network Analysis in Python
Complex and Social Network Analysis in Python
 
SSA slides
SSA slidesSSA slides
SSA slides
 
Social Network Analysis
Social Network AnalysisSocial Network Analysis
Social Network Analysis
 
Lesson 26: The Fundamental Theorem of Calculus (Section 4 version)
Lesson 26: The Fundamental Theorem of Calculus (Section 4 version)Lesson 26: The Fundamental Theorem of Calculus (Section 4 version)
Lesson 26: The Fundamental Theorem of Calculus (Section 4 version)
 
WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...
WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...
WE4.L09 - MEAN-SHIFT AND HIERARCHICAL CLUSTERING FOR TEXTURED POLARIMETRIC SA...
 
16 fft
16 fft16 fft
16 fft
 
Lesson 27: Evaluating Definite Integrals
Lesson 27: Evaluating Definite IntegralsLesson 27: Evaluating Definite Integrals
Lesson 27: Evaluating Definite Integrals
 
An introduction to quantum stochastic calculus
An introduction to quantum stochastic calculusAn introduction to quantum stochastic calculus
An introduction to quantum stochastic calculus
 
TunUp final presentation
TunUp final presentationTunUp final presentation
TunUp final presentation
 
Lesson 27: Evaluating Definite Integrals
Lesson 27: Evaluating Definite IntegralsLesson 27: Evaluating Definite Integrals
Lesson 27: Evaluating Definite Integrals
 
Lesson 26: The Fundamental Theorem of Calculus (Section 10 version)
Lesson 26: The Fundamental Theorem of Calculus (Section 10 version)Lesson 26: The Fundamental Theorem of Calculus (Section 10 version)
Lesson 26: The Fundamental Theorem of Calculus (Section 10 version)
 
Taylor problem
Taylor problemTaylor problem
Taylor problem
 
Lesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of CalculusLesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of Calculus
 
Lesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of CalculusLesson 28: The Fundamental Theorem of Calculus
Lesson 28: The Fundamental Theorem of Calculus
 
C2.6
C2.6C2.6
C2.6
 
The Probability that a Matrix of Integers Is Diagonalizable
The Probability that a Matrix of Integers Is DiagonalizableThe Probability that a Matrix of Integers Is Diagonalizable
The Probability that a Matrix of Integers Is Diagonalizable
 
Divergence center-based clustering and their applications
Divergence center-based clustering and their applicationsDivergence center-based clustering and their applications
Divergence center-based clustering and their applications
 
Information-theoretic clustering with applications
Information-theoretic clustering  with applicationsInformation-theoretic clustering  with applications
Information-theoretic clustering with applications
 
Approximate Tree Kernels
Approximate Tree KernelsApproximate Tree Kernels
Approximate Tree Kernels
 
Triangle counting handout
Triangle counting handoutTriangle counting handout
Triangle counting handout
 

Kürzlich hochgeladen

The basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptxThe basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptx
heathfieldcps1
 

Kürzlich hochgeladen (20)

Sociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning ExhibitSociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning Exhibit
 
FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024
 
Towards a code of practice for AI in AT.pptx
Towards a code of practice for AI in AT.pptxTowards a code of practice for AI in AT.pptx
Towards a code of practice for AI in AT.pptx
 
Application orientated numerical on hev.ppt
Application orientated numerical on hev.pptApplication orientated numerical on hev.ppt
Application orientated numerical on hev.ppt
 
Basic Civil Engineering first year Notes- Chapter 4 Building.pptx
Basic Civil Engineering first year Notes- Chapter 4 Building.pptxBasic Civil Engineering first year Notes- Chapter 4 Building.pptx
Basic Civil Engineering first year Notes- Chapter 4 Building.pptx
 
Understanding Accommodations and Modifications
Understanding  Accommodations and ModificationsUnderstanding  Accommodations and Modifications
Understanding Accommodations and Modifications
 
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
 
Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...
Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...
Sensory_Experience_and_Emotional_Resonance_in_Gabriel_Okaras_The_Piano_and_Th...
 
Fostering Friendships - Enhancing Social Bonds in the Classroom
Fostering Friendships - Enhancing Social Bonds  in the ClassroomFostering Friendships - Enhancing Social Bonds  in the Classroom
Fostering Friendships - Enhancing Social Bonds in the Classroom
 
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
 
Plant propagation: Sexual and Asexual propapagation.pptx
Plant propagation: Sexual and Asexual propapagation.pptxPlant propagation: Sexual and Asexual propapagation.pptx
Plant propagation: Sexual and Asexual propapagation.pptx
 
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptxExploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
Exploring_the_Narrative_Style_of_Amitav_Ghoshs_Gun_Island.pptx
 
The basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptxThe basics of sentences session 3pptx.pptx
The basics of sentences session 3pptx.pptx
 
Holdier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdfHoldier Curriculum Vitae (April 2024).pdf
Holdier Curriculum Vitae (April 2024).pdf
 
Single or Multiple melodic lines structure
Single or Multiple melodic lines structureSingle or Multiple melodic lines structure
Single or Multiple melodic lines structure
 
Unit 3 Emotional Intelligence and Spiritual Intelligence.pdf
Unit 3 Emotional Intelligence and Spiritual Intelligence.pdfUnit 3 Emotional Intelligence and Spiritual Intelligence.pdf
Unit 3 Emotional Intelligence and Spiritual Intelligence.pdf
 
Google Gemini An AI Revolution in Education.pptx
Google Gemini An AI Revolution in Education.pptxGoogle Gemini An AI Revolution in Education.pptx
Google Gemini An AI Revolution in Education.pptx
 
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptxOn_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
On_Translating_a_Tamil_Poem_by_A_K_Ramanujan.pptx
 
General Principles of Intellectual Property: Concepts of Intellectual Proper...
General Principles of Intellectual Property: Concepts of Intellectual  Proper...General Principles of Intellectual Property: Concepts of Intellectual  Proper...
General Principles of Intellectual Property: Concepts of Intellectual Proper...
 
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
 

nips勉強会_Toward Property-Based Classification of Clustering Paradigms

  • 2.   1     Twitter ◦ 
  • 3.     Properties of Clustering Functions   A taxonomy of k-clustering fucntions    
  • 4.     Properties of Clustering Functions   A taxonomy of k-clustering fucntions    
  • 5. Clustering: ↓ Clustering Ad-hoc ○Clustering ○ Clustering property
  • 6. Clustering clustering
  • 7.   A Impossibility Theorem for Clustering ◦  Jon Kleinberg, NIPS 2002   Measures of Clustering Quality: A Working Set of Axioms for Clustering ◦  M.Ackerman and S.Ben-David, NIPS 2008   Characterization of Linkage-based Clustering. ◦  M.Ackerman and S.Ben-David, COLT 2010
  • 8. X: + d : X × X →R ,d ( x, x ) = 0(∀x ∈ X ) X,d ) ( clustering ( X,d,k ) F clustering € C = F ( X,d,k ) € ⎛ € ⎞ {C1,C2 ,Ck }⎜ Ci = X,1 ≤ k ≤ X ⎟ ⎝ i ⎠
  • 9. general clustering function F Input: F ( X,d ) ⎛ ⎞ Output: C = {C1,C2 ,Ck }⎜ Ci = X,1 ≤ k ≤ X ⎟ ⎝ i ⎠ € k-clustering function F € Input: F ( X,d,k ), (1 ≤ k ≤ X ) ⎛ ⎞ Output: = {C1,C2 ,Ck }⎜ Ci = X,1 ≤ k ≤ X ⎟ C ⎝ i ⎠ € €
  • 10.     Properties of Clustering Functions   A taxonomy of k-clustering fucntions    
  • 11. iso. invariance scale invariance order invariance outer consistency cluster inner consistency cluster k-rich k inner rich richness outer rich threshold rich locality clustering refinement-confined k clustering
  • 12. clustering φ : X → X ʹ′ x, y ∈ X,d ( x, y ) = dʹ′(φ (x), φ (y)) F ( X,d,k ),F ( X ʹ′, dʹ′,k ) : isomorphic(∀k) x, y : same → φ (x), φ (y) : same
  • 14. clustering x, y ∈ X, d ( x, y ) = cdʹ′( x, y ) →F ( X,d,k ) = F ( X ʹ′, dʹ′,k )
  • 15. 2 clustering x1, x 2 , x 3 , x 4 ∈ X, d ( x1, x 2 ) < d ( x 3 , x 4 ), dʹ′( x1, x 2 ) < dʹ′( x 3 , x 4 ) →F ( X,d,k ) = F ( X ʹ′, dʹ′,k ) €
  • 16. 2 clustering (Single-linkage clustering 0 1 4 9 10 12 15 19 20
  • 17. clustering cluster cluster C’
  • 18. clustering cluster cluster C = F(X,d,k), Cʹ′ ⊆ C F(Cʹ′,d,| Cʹ′ |)= Cʹ′
  • 19. cluster cluster clustering cluster cluster d(x,y) d’(x,y) d(x,y) d’(x,y)
  • 20. cluster cluster clustering cluster cluster dʹ′ : (C,d ) − consistent x, y : same → dʹ′( x, y ) ≤ d ( x, y ) x, y : different → dʹ′( x, y ) ≥ d ( x, y )
  • 21. cluster clustering cluster d(x,y) d’(x,y)
  • 22. cluster clustering cluster dʹ′ : (C,d) − outerconsistent x, y : same → dʹ′( x, y ) = d ( x, y ) x, y : different → dʹ′( x, y ) ≥ d ( x, y )
  • 23. cluster clustering cluster d(x,y) d’(x,y)
  • 24. cluster clustering cluster dʹ′ : (C,d) − innterconsistent x, y : same → dʹ′( x, y ) ≤ d ( x, y ) x, y : different → dʹ′( x, y ) = d ( x, y )
  • 25. clustering any : X1, X 2  X k X ʹ′ = { X1, X 2  X k } →∃d : F ( X ʹ′,d,k ) = { X1, X 2  X k }
  • 26. clustering ※ clustering any : (X1,d1 ),(X 2 ,d2 )(X k ,dk ) ⎛ k ⎞ →∃d : F ⎜ X i , d,k ⎟ = { X1, X 2  X k } ˆ ˆ ⎝ i=1 ⎠ ˆ d : entends − d (i ≤ k) i
  • 27. clustering ※ clustering ( X,d), X = { X1, X 2  X k } ˆ ˆ ( →∃d : d ( a,b) = d ( a,b) a ∈ X i ,b ∈ X j ,i ≠ j ) ⎛ k ⎞ F ⎜  X i , d,k ⎟ = { X1, X 2  X k } ˆ ⎝ i=1 ⎠
  • 28. clustering ∃a < b x, y : same →d(x, y) ≤ a, x, y : different →d(x, y) ≥ b F ( X,d, C ) = C
  • 29. k≦k’ F(X,d,k’) F(X,d,k) refinement 1 ≤ k ≤ kʹ′ ≤ X , O( F ( X,d,k')) ≥ O( F ( X,d,k )) €
  • 30. k≦k’ F(X,d,k’) F(X,d,k) refinement
  • 31. iso. invariance scale invariance order invariance outer consistency cluster inner consistency cluster k-rich k inner rich richness outer rich threshold rich locality clustering refinement-confined k clustering
  • 32.   Clustering Kleinberg clustering Scale-Invariance Richness Consistency
  • 33.   single linkage clustering stop condition   Consistency + Richness: only link if distance is less than r ◦  clustering cluster   Consistency + SI: stop when you have k connected components ◦  clustering / clustering   Richness + SI: if x is the diameter of the graph, only add edges with weight βx ◦  cluster / clustering
  • 34.     Properties of Clustering Functions   A taxonomy of k-clustering fucntions    
  • 35. outer consistent inner consistent scale invariant order invariant threshold rich iso. invariant refinement inner rich out rich k-rich local Singe Linkage ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ Average Linkage ○ × ○ ○ × ○ ○ ○ ○ ○ ○ Complete Linkage ○ × ○ ○ ○ ○ ○ ○ ○ ○ ○ k-median ○ × ○ × × ○ ○ ○ ○ ○ ○ k-means ○ × ○ × × ○ ○ ○ ○ ○ ○ Min sum ○ ○ ○ × × ○ ○ ○ ○ ○ ○ Ratio cut × ○ × × × ○ ○ ○ ○ ○ ○ Nomalize cut × × × × × ○ ○ ○ ○ ○ ○
  • 36. outer consistent inner consistent scale invariant order invariant threshold rich iso. invariant refinement inner rich out rich k-rich local Singe Linkage ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ Average Linkage ○ × ○ ○ × ○ ○ ○ ○ ○ ○ Complete Linkage ○ × ○ ○ ○ ○ ○ ○ ○ ○ ○ k-median ○ × ○ × × ○ ○ ○ ○ ○ ○ k-means ○ × ○ × × ○ ○ ○ ○ ○ ○ Min sum ○ ○ ○ × × ○ ○ ○ ○ ○ ○ Ratio cut × ○ × × × ○ ○ ○ ○ ○ ○ Nomalize cut × × × × × ○ ○ ○ ○ ○ ○
  • 37. outer consistent inner consistent scale invariant order invariant threshold rich iso. invariant refinement inner rich out rich k-rich local Singe Linkage ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ Average Linkage ○ × ○ ○ × ○ ○ ○ ○ ○ ○ Complete Linkage ○ × ○ ○ ○ ○ ○ ○ ○ ○ ○ k-median ○ × ○ × × ○ ○ ○ ○ ○ ○ k-means ○ × ○ × × ○ ○ ○ ○ ○ ○ Min sum ○ ○ ○ × × ○ ○ ○ ○ ○ ○ Ratio cut × ○ × × × ○ ○ ○ ○ ○ ○ Nomalize cut × × × × × ○ ○ ○ ○ ○ ○
  • 38. outer consistent inner consistent scale invariant order invariant threshold rich iso. invariant refinement inner rich out rich k-rich local Singe Linkage ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ Average Linkage ○ × ○ ○ × ○ ○ ○ ○ ○ ○ Complete Linkage ○ × ○ ○ ○ ○ ○ ○ ○ ○ ○ k-median ○ × ○ × × ○ ○ ○ ○ ○ ○ k-means ○ × ○ × × ○ ○ ○ ○ ○ ○ Min sum ○ ○ ○ × × ○ ○ ○ ○ ○ ○ Ratio cut × ○ × × × ○ ○ ○ ○ ○ ○ Nomalize cut × × × × × ○ ○ ○ ○ ○ ○
  • 39. outer consistent inner consistent scale invariant order invariant threshold rich iso. invariant refinement inner rich out rich k-rich local Singe Linkage ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ Average Linkage ○ × ○ ○ × ○ ○ ○ ○ ○ ○ Complete Linkage clustering○ ○ × ○ ○ ○ ○ ○ ○ ○ ○ scale invariance k-median ○ × ○ : × × natural ○ ○ ○ ○ ○ ○ isomorphism variance :natural k-means ○ × ○ × × ○ ○ ○ ○ ○ ○ threshold richness Min sum ○ ○ ○ × / × ○ ○ ○ ○ ○ ○ clustering Ratio cut × ○ × × × ○ ○ ○ ○ ○ ○ Nomalize cut × × × × × ○ ○ ○ ○ ○ ○
  • 40.     Properties of Clustering Functions   A taxonomy of k-clustering fucntions    
  • 41.   Invariance properties   Consistency properties (C,d) − nice var iant [ ] [ P F ( X, dʹ′, C ) = C ≥ P F ( X,d, C ) = C ]   Richness properties ∀ε > 0 € ∃d : P ( F ( X,d,k ) = C ) ≥1 − ε   Locality [ ] P F ( X ʹ′,d / X ʹ′, Cʹ′ ) = Cʹ′ € = [ P Cʹ′ ⊆ C F ( X,d, j ) = CandC / X ʹ′isak − clustering ] [ ] P ∃C1,C2 Ck s.t.Ci = X ʹ′ F ( X,d, j ) = C ≠ 0
  • 42.   k-means
  • 43.   k-means
  • 44. Random Centroids Lloyd Furthest Centroids Lloyd ( maximizemin1≤ j ≤i−1 d c j ,c i ) €
  • 45. Clustering Algorithm outer consistent scale invariant threshold rich iso. invariant outer rich k-rich local Optimal k-means ○ ○ ○ ○ ○ ○ ○ Random Centroid Lloyd × × × ○ ○ ○ ○ Furthest Centroid Lloyd × × ○ ○ ○ ○ ○
  • 46. Clustering Algorithm outer consistent scale invariant threshold rich iso. invariant outer rich k-rich local Optimal k-means ○ ○ ○ ○ ○ ○ ○ Random Centroid Lloyd × × × ○ ○ ○ ○ Furthest Centroid Lloyd × × ○ ○ ○ ○ ○ threshold richness Furthest Centroid Lloyd Random Centroid Lloyd
  • 47.   Kleinberg clustering Scale-Invariance Richness Consistency
  • 48.   clustering Scale-Invariance Richness Outer-Consistency
  • 49.     Properties of Clustering Functions   A taxonomy of k-clustering fucntions    
  • 50. Clustering Function property clustering axioms scale-invariance, isomorphism-invariance, threshold richness Kleinberg
  • 51.   Supervised Clustering ◦  2008 clustering ◦  clustering merge/ split   Efficient Robust Feature Selection via Joint L2,1-Norms Minimization ◦  Bio Informatics Feature Selection ◦  L1,2-norm SVM Feature