Missing data intuition

Here's an intuition check! When handling missing data, you have three options: keep, replace, and remove.

You've been looking at numeric columns, but what about a non-numeric column? How would you handle missing data in the column person_home_ownership which has string values?

The object ownership_table has already been created to show how many records occur in each unique value of person_home_ownership with the following code:

# Count the number of records for each unique value
cr_loan['person_home_ownership'].value_counts()

ownership_table and cr_loan are already loaded in the workspace.

Possible answers

Remove the records, they aren't going to be useful for prediction anyway.

Replace the data with the value Other.

Replace the data with the most frequently occurring value, Rent.

Keep the missing values because we don't know what to choose.

Exercise

Missing data intuition

Instructions

Possible answers