Another extremely important part of machine learning methods are the time-series forecast, which is the same as the models before, but instead of depends on category and number features, it depends on past values ordered by the time, i.e., the actual values depends on the nearer past values.
For this purpose, we applied less theory about time series methods and more business models in time series (econometrics). Also, the main target is make a business analysis than a scientifical one, applying facebook autoML algorithm Prophet
Personal pages: