Which of the following statements regarding clustering and anomaly detection is INCORRECT?

A$.$ Sometimes we can apply anomaly detection first before running a clustering algorithm and get better clustering result.

B$.$ The result of clustering can shed light into finding anomalies.

C$.$ Both clustering and anomaly detection typically have no training data to begin with.

D$.$ If some data is hard to cluster, then it must be hard to find anomalies in it$.$