Clustering German Credit Data using K-Means

Full Service Business Data Solutions

In this discussion board, we will try to use k-means unsupervised learning algorithm for German credit card data. We observed two main problem with this assignment as the questions asked us to use only three columns and we are losing information that could be very much helpful for the k-means. Second, the cross-validation question is not valid for k-means, but it is meant for KNN therefore we cannot calculate the cross-validation with k = 5. However, we attempt to show the comparison between the original label and the predicted class using k-means.