comparing results of Kmeans algorithm with Database to find out The precision of algorithm

Hi, I have 2000 articles(2000 .txt files) from 20 subjects(20 Folders). it's my Database.

I clustered them by Kmeans Algorithm.("idx" parametr in Kmeans , shows me Each article belongs to which cluster)

Now , How can i compare Kmeans Result With Database to find out The precision of algorithm?

it's hard to use "Eye" for 2000 files!

Answer 1

[Image Analyst]

This is typically done with a "confusion matrix" which is a table of N classes by N classes that shows you what class a sample got classified as, versus what it's "True" class is. Ideal classification would yield a confusion matrix with numbers only along the diagonal. The more off-diagonal it becomes, the less accurate your classification algorithm is.

You can also use ROC curves http://en.wikipedia.org/wiki/Receiver_operating_characteristic which is a plot of true positives vs. false negatives. ROC curves are especially used in clinical studies.

[Isay]

Do you have matlab code about that ? like as ROC ?

[Image Analyst]

Sorry, no. Check the File Exchange.