This discussion is often referred to as the ontology of the data in the literature. How is the data actually broken down and used? What are the important classes? Can you describe the class distribution and types? These are topics that would be explored in a report focused on the ontology. As you delve deeper into this part of the science, you'll find that there are also a bunch of topics such as anomaly detection and balancing of the data. In practice, it's hard to come by a dataset that is balanced from the onset. Typically, you'll need to do preparation with the data. We'll touch on both topics in this chapter.

Now you understand some of the common types of data types that we will work with this in this field. I am going to suggest a few links to get more educated on other possible data types or ways of using the data with the following links:

• Data types with Python: https://developer.rhino3d.com/guides/rhinopython/python-datatypes/
• Data types within TensorFlow: https://www.tensorflow.org/versions/r1.2/programmers_guide/dims_types