Preparing Tabular Data for Neural Networks

This post could also be called "everything I learned during the first year of my PhD while working with a massive electronic health record data set." Here, I'll overview the difference between categorical and quantitative variables, and I'll explain how to prepare tabular data for a neural network model, including: Best of all, broadly-applicable code is provided to do all of the above on your tabular data! A categorical variable is a variable that takes on different category names. Quantitative variables have numerical values that usually represent the result of a measurement. If your variable is the NAME of a laboratory test, then that's a nominal (categorical) variable because the values it can take on are different names of different lab tests, e.g.