Scribd
Upload
361 views3 pages

Correction Exercices DIANA Et AGnes

The document describes the AGNES and DIANA hierarchical clustering algorithms. It provides an example of applying each algorithm to a dataset with 5 objects (Xav, Yves, Ziad, Tania, Ute). For AGNES, it shows the successive merges that are made at each step based on minimum distance. For DIANA, it shows how objects are successively split off from the cluster based on maximum distance at each step.

Uploaded by

Marwen Guesmi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
361 views3 pages

Correction Exercices DIANA Et AGnes

The document describes the AGNES and DIANA hierarchical clustering algorithms. It provides an example of applying each algorithm to a dataset with 5 objects (Xav, Yves, Ziad, Tania, Ute). For AGNES, it shows the successive merges that are made at each step based on minimum distance. For DIANA, it shows how objects are successively split off from the cluster based on maximum distance at each step.

Uploaded by

Marwen Guesmi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Algorithme AGNES :

Dans le cours, nous avons suivi le saut minimal comme méthode d’aggrégation

Méthode d’agrégation : Saut moyen

1ère fusion : la distance minimale dans notre dataset est entre Tania et Utes =1.5

Donc la 1ere fusion est : {Tania, Utes}

Mise à jour de la table de ressemblance

Xav Yves Ziad {tania, utes}


Xav 0
yves 4.5 0
Ziad 2 3.5 0
{tania, utes} 4.5 5.75 3.5 0

2eme fusion : la distance minimale dans notre table de ressemblance est entre Xav et Ziad =2

Donc la 2eme fusion est : {Xav, Ziad}

Mise à jour de la table de ressemblance

{Xav, Ziad} Yves {tania, utes}


{Xav, Ziad} 0
yves 4 0
{tania, utes} 4 5.75 0

3eme fusion : la distance minimale dans notre table de ressemblance est 4 :

1. entre {Xav , Ziad } et Yves


2. entre {Xav , Ziad} et {Tania, Utes}

Donc la aléatoirement nous allons choisir {Tania, Utes}, donc la nouvelle fusion est :

{ Xav , Ziad , Tania, Utes},

Mise à jour de la table de ressemblance

{Xav, Ziad, Tania, Utes} Yves


{Xav, Ziad, tania, utes} 0
yves 4.88 0

4eme fusion : {Xav, Ziad, tania, utes, yves}


Algorithme DIANA :
C= {Xav, Yves, Ziad, Tania, Ute}

1ere étape:
AC; B ={}

A= {Xav, Yves, Ziad, Tania, Ute}

Distance (Xav, A) = (distance (Xav, Yves) + distance (Xav, Ziad) + distance (Xav, Tania) + distance (Xav,
Ute)) / 4 = (4.5 +2 +5 +4) / 4 = 15.5/4

Distance (Yves, A) = (distance (Yves, Xav) + distance (Yves, Ziad) + distance (Yves, Tania) + distance
(Yves, Ute)) / 4 = (4.5 +3.5 + 6 + 5.5) / 4 = 19.5/4

Distance (Ziad, A) = (distance (Ziad, Xav) + distance (Ziad, Yves) + distance (Ziad, Tania) + distance
(Ziad, Ute)) / 4 = (2 +3.5 + 4.5 + 2.5) / 4 = 12.5/4

Distance (Tania, A) = (distance (Tania, Xav) + distance (Tania, Yves) + distance (Tania, Ziad) + distance
(Tania, Ute)) / 4 = (5 + 6 + 4.5 + 1.5) / 4 = 17/4

Distance (Ute, A) = (distance (Ute, Xav) + distance (Ute, Yves) + distance (Ute, Ziad) + distance (Ute,
Tania)) / 4 = (4 + 5.5 + 2.5 + 1.5) / 4 = 13.5/4

 Yves est le plus distant au cluster A


 A= {Xav, Ziad, Tania, Ute}
 B = {Yves}
1. a. Distance (Xav, A) = (distance (Xav, Ziad) + distance (Xav, Tania) + distance (Xav, Ute)) / 3 =
(2 +5 +4) / 3 = 11/3
b. Distance (Xav, B) = distance (Xav, Yves) = 4.5 > 11/3
2. a. Distance (Ziad, A) = (distance (Ziad, Xav) + distance (Ziad, Tania) + distance (Ziad, Ute)) / 3 =
(2 + 4.5 + 2.5) / 3 = 9/3 =3

b. Distance (Ziad, B) = distance (Ziad, Yves) = 3.5> 3

3. a. Distance (Tania, A) = (distance (Tania, Xav) + distance (Tania, Ziad) + distance (Tania, Ute)) /
3 = (5 + 4.5 + 1.5) / 3 = 11/3

b. Distance (Tania, B) = distance (Tania, Yves) = 6>11/3

4. a. Distance (Ute, A) = (distance (Ute, Xav) + distance (Ute, Yves) + distance (Ute, Ziad) +
distance (Ute, Tania)) / 3 = (4 + 2.5 + 1.5) / 3 = 8/3

b. Distance (Ute, B) = distance (Ute, Yves) = 5.5 >8/3

 A= {Xav, Ziad, Tania, Ute}


 B = {Yves}

2ème étape:
C = {Xav, Ziad, Tania, Ute}
AC; B ={}

A= {Xav, Ziad, Tania, Ute}

Distance (Xav, A) = (distance (Xav, Ziad) + distance (Xav, Tania) + distance (Xav, Ute)) / 3 = (2 +5 +4) /
3 = 11/3

Distance (Ziad, A) = (distance (Ziad, Xav) + distance (Ziad, Tania) + distance (Ziad, Ute)) / 3 = (2 + 4.5 +
2.5) / 3 = 9/3 =3

Distance (Tania, A) = (distance (Tania, Xav) + distance (Tania, Ziad) + distance (Tania, Ute)) / 3 = (5 +
4.5 + 1.5) / 3 = 11/3

Distance (Ute, A) = (distance (Ute, Xav) + distance (Ute, Yves) + distance (Ute, Ziad) + distance (Ute,
Tania)) / 3 = (4 + 2.5 + 1.5) / 3 = 8/3

 Tania et Xav sont les plus distants au cluster A, on va faire sortir par exemple Tania
 A= {Xav, Ziad, Ute}
 B = { Tania }
1. a. Distance (Xav, A) = (distance (Xav, Ziad) + distance (Xav, Ute)) / 2 = (2 +4) / 2 = 3
b. Distance (Xav, B) = distance (Xav, Tania) = 5 > 3
2. a. Distance (Ziad, A) = (distance (Ziad, Xav) + distance (Ziad, Ute))/2 = (2 + 2.5)/2 = 4.5/2=2.25

b. Distance (Ziad, B) = distance (Ziad, Tania) = 4.5> 2.25

4. a. Distance (Ute, A) = (distance (Ute, Xav) + distance (Ute, Ziad)) / 2 = (4 + 2.5)/2= 6.5/2 =3.25

b. Distance (Ute, B) = distance (Ute, Tania) = 1.5 < 3.25

Ute est plus proche au cluster B, donc Ute passe de A à B

 A= {Xav, Ziad}
 B = {Tania, Ute}

3ème étape:
distance (Xav, Ziad) > distance (Tania, Ute)

donc C = {Xav, Ziad} (elle présente la plus grande dissimilarité entre 2 objets)

 A= {Xav }
 B = {Ziad}

4ème étape:
C= {Tania, Ute}

 A= {Tania}
 B = {Ute}

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy