en

0.25

0.5

0.75

1.25

1.5

1.75

2

# α-Clusterable Sets

Published on Oct 03, 20113417 Views

In spite of the increasing interest into clustering research within the last decades, a unified clustering theory that is independent of a particular algorithm, or underlying the data structure and

#### Related categories

#### Chapter list

α-Clusterable Sets00:00

Outline00:13

What is Clustering?01:01

Motivation01:42

Contribution02:15

Window Density Function (WDF)02:41

Example plots of WDF (1)03:34

Example plots of WDF (2)04:22

Basic Notions05:04

Example of α Clusterable Set05:57

Basic Notions06:46

α-Clustering Function Properties (1)07:23

α-Clustering Function Properties (2)07:36

α-Clustering Function Properties (3)07:54

α-Clustering Function Properties (4)08:12

Proposed Algorithm08:33

Steps of the algorithm "PSO α-Cl" (1)09:26

Steps of the algorithm "PSO α-Cl" (2)09:28

Steps of the algorithm "PSO α-Cl" (3)09:34

Steps of the algorithm "PSO α-Cl" (4)09:42

Steps of the algorithm "PSO α-Cl" (5)09:48

Steps of the algorithm "PSO α-Cl" (6)09:52

Steps of the algorithm "PSO α-Cl" (7)09:59

Steps of the algorithm "PSO α-Cl" (8)10:05

Step 1 of the algorithm10:17

Step 2: Perform the PSO algorithm to return the centre z of an α-Clusterable Set (1)10:22

Step 2: Perform the PSO algorithm to return the centre z of an α-Clusterable Set (2)11:34

Step 2: Perform the PSO algorithm to return the centre z of an α-Clusterable Set (3)11:45

Step 2: Perform the PSO algorithm to return the centre z of an α-Clusterable Set (4)11:54

Steps 3 and 4 (1)12:03

Steps 3 and 4 (2)12:08

Repeat the previous Steps 1-4 for the remaining dataset (1)12:25

Repeat the previous Steps 1-4 for the remaining dataset (2)12:35

Steps 2, 3 and 4 (1)12:37

Steps 2, 3 and 4 (2)12:39

After 16th iterations, the algorithm attains to detect the α-Clusterable Sets (1)12:40

After 16th iterations, the algorithm attains to detect the α-Clusterable Sets (2)12:53

Goals of the Experiments13:09

Entropy and Purity vs Window Size α (1)13:34

Entropy and Purity vs Window Size α (2)14:31

Entropy and Purity vs Window Size α (3)14:45

Performance of PSO α-Cl algorithm VS other clustering algorithms14:52

Scalability of the "PSO α-Cl" algorithm (1)15:56

Scalability of the "PSO α-Cl" algorithm (2)16:39

Concluding Remarks - Future Work17:01

Thank you for your attention...17:33