hi, my research is based on intrusion detection detection using KNN and Bayes on first stage then k means clustering after. I am concerned with how im going to "un-label" my instances when they come out of the supervised stage. what i have in mind is that i can just discard the label i.e classes and just take the features as they are, but my question is since to categorize that instance to one of the clusters, i need to calculate distance so what values am i going to use from the unlabelled instance? i hope this makes sense
I suggest you follow: First simple example: http://mnemstudio.org/clustering-k-means-example-1.htm
https://www.edureka.co/blog/k-means-clustering/
https://towardsdatascience.com/k-means-clustering-algorithm-applications-evaluation-methods-and-drawbacks-aa03e644b48a
Apply the agglomeration classification and calculate 'KNN' by using Bayes naive
- Badges
- Science topic
- Similar topics
- Comment