Let There Be Light! Using Machine Learning to Replace a Light Switch

This study examines the possibilities that come with using existing smart home equipment to turn on and off devices based on changes in the state of other devices.  The solution was an exploration into Auto-ML as it is designed for end consumers that will not be able to do proper model selection.  The project found a lot of very exciting insights into relationships in a home no one would guess.  It also found in determining if a device will change within a window of time or when events occur calculating how many minutes until a change in another device was very accurate even against cross validation.   This is likely going to be a burgeoning field for data science over the next decade.