I create 4 clusters by using k-mean cluster analysis. After that, I try to classify this target variable with random forest in order to confirm if the clusters are properly grouped. Is it possible?
Generally speaking it is possible. Somemore information would be helpful: How do you plan to test/train? Do you want to use k-fold cross validation, or split the data? Are you going to use exactly the same variables as predictors in the Random Forest, that u used for the clusteranalysis? In this case you propably need to be careful with your interpretation since it is circular by nature.