Skip to content

TahaTabatabaei/Image-clustering

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Image-clustering

simple image clustring (university assignment)

In the first phase, attempt is to try 3 diffrent clustering methods over the ORL dataset: K-means, DBSACN & Agglomerative (Single Link, Complete Link, Group Average)

also, there is a basic implementation of "Rand Index" metric from scratch for result evaluating in Rand_index.py

In the second phase, we find a way to estimate epsilon parameter, in order to improve DBSACN clustering. we observe the average distance of a data to its 5 nearest nighboors. this average lead us to find a reasonable epsilon. We assume that this average distance is a good estimation for density distribution of dataset.