I am doing k-means clustering on a data and using the FeatureImpCluster for getting the important features. This is great and I understand that the idea is to randomize a specific feature and see how it impacts the misclassification error while predicting. So, I want to know more about the model that this function uses for prediction? In my case, does it redo kmeans clustering and checks how the classification is different for the data points or does it use another ML model such as random forest to predict the classification?
Hi,
I am doing k-means clustering on a data and using the FeatureImpCluster for getting the important features. This is great and I understand that the idea is to randomize a specific feature and see how it impacts the misclassification error while predicting. So, I want to know more about the model that this function uses for prediction? In my case, does it redo kmeans clustering and checks how the classification is different for the data points or does it use another ML model such as random forest to predict the classification?
Please let me know.