PSO Clustering

16 views (last 30 days)
Amishi
Amishi on 24 Feb 2012
Hi, I am implementing PSO clustering technique.
I have a query regarding how a particle should be updated
The algorithm is as follows :
1. Initialize each particle with random cluster centeroid
2. Do
2.1. For each Particle
2.1.1. For Each Data Vector Zp in Data set
2.1.2. Calculate D(mij,Zp) for all cluster centeroids.
2.1.3. Assign Zp to a Cluster Ci which has minimum Euclidean distance
2.3 Calculate Fitness of Particle using the objective function
2.4 Update the Global Best and Local Best
2.5. Update the Particle Velocity and Location using Equation 1 and 2.
3. While termination Criteria is satisfied
vid = w × vid + c1 × rand1( ) × (pid xid) +c2 × rand2( ) × (pgd xid) -eq 1
xid = xid + vid -eq 2

Answers (2)

Gnaneswar Nadh satapathi
Gnaneswar Nadh satapathi on 4 Jun 2015
Please go through the below code, it might help you
https://github.com/iralabdisco/pso-clustering/blob/123c8a658fb914cb1a5d797ec9b7b495a08d325c/pso_clustering.m

moqueed ahmed
moqueed ahmed on 24 Apr 2015
if i am using fitness function then getting only one fit value
  1 Comment
Walter Roberson
Walter Roberson on 4 May 2015
You should create a new question for this, showing what you have tried.

Sign in to comment.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!