This step will be the first thing to think of after getting a new dataset from the customer, because there will be missing/incorrect data in nearly every dataset. In the next chapters, you will see that some learning algorithms are able to deal with missing values and others need you to handle missing data. During this example, we are going to use the random forest classifier from scikit-learn, which requires separate handling of missing data.

There are different approaches that you can use to handle missing data.