Ahmad is a data scientist for a company that develops mobile apps. His team is working on a project to recommend mobile apps based on user preferences. The goal of this machinelearning project is to build a recommendation engine that recommends apps to users.

Ahmad needs to predict future behavior based on past data. He builds and cleans a data set of over $100,000$ data points. He has three columns of user data: userID, appID, and rating.

In which step of the data science methodology is Ahmad working?

Data representation and transformation

Deploy data models

Business understanding

Data exploration and preparation

Train data models

Data visualization and presentation