I submitted too fast. There is a similar thread mentioning that the similar error can be thrown with a Shared cluster.

https://community.databricks.com/t5/data-engineering/constructor-public-org-apache-spark-ml-feature-...