MLS C1 W2 Defining features in regression

This is a question to better understand how to identify and define features.

In the example for predicting housing prices, how would I include features such as the state or zip code? A house in California will be much pricier than a similar house in Kentucky. Even within California, the zip code of the house might influence the price. How do we define such features? We can’t use the alphanumeric state as is… and zip codes are just random numbers and their values don’t influence the price (unlike the frontage or total area of the house). So, if I wanted to include such features how would we go about it?

Geographic regions can be identified using the latitude and longitude (in decimal degrees) for the centroid of the area, as two separate features.