machine learning - How to derive cluster properties -


i have clustered ~40000 points 79 clusters. each point vector of 18 features. want 'derive' characteristics of each cluster - prominent features/characteristics of clusters. there machine-learning algorithms derive this?

if confident clusters meaningful particular needs, view classification problem.

one option apply feature selection algorithm rank features. use recursive feature elimination identify subset of features predictive cluster labels.

another option interpreting clusters building decision tree. decision trees can see features used best separate classes (clusters in case). use ensemble random forest , ask feature importance scores.


Comments

Popular posts from this blog

get url and add instance to a model with prefilled foreign key :django admin -

css - Make div keyboard-scrollable in jQuery Mobile? -

ruby on rails - Seeing duplicate requests handled with Unicorn -