My model uses pre-processed data to predict if a customer is a private or a non-private customer. The pre-processing-step is using steps like feature_column.bucketized_column(…)
