I am pretty confused about the following two problems: I have a 15-dimensional dataset that should be used to group how many types of attacks a dataset contains.
1. Now I have already grouped my data set into 5 clusters (5 attacks). Does anyone know how I can indicate which cluster is the attack? (how to designate clusters is not just "cluster 1, cluster 2 ...")
2. In a controlled classification, we have a set of training materials and a data set for testing, and testing is carried out using a classifier built from a training data set. My question is whether the same approach can be used for clustering. How to build a model with a clustering algorithm and then automatically classify a new instance into a specific cluster? Is it possible?
source
share