You want to identify the breaker switch for a particular outlet.
This problem inspired me to write an article on Medium one day as I was thinking about the electric wiring at home. As you’ll see, there are commonalities between decidely challenging tasks in Machine Learning and the problems you might encounter day to day.
You have say ten switches and, “Hey,” you might decide, “why not test each switch one at a time to see if I toggled the outlet?” Not so bad. But if you have any more than ten, or you’re otherwise optimization focused in general, then you’d have scalability in mind. In particular, you’d be thinking about mathematical properties like convergence and success rate. This is after all a problem of global optimization, even though you might not feel like breaking out your diploma for a measly switch.
So, you panic and the ten switches suddenly look like ten-thousand. You find an electrician online but they don’t answer your calls. Big deal. Why not try your hand? Can you think of a faster way to find the right breaker switch so you can plug in that new toaster without taking the whole house down?
Solution: optimize your search space by considering its constraints. We’re checking residential breaker switches, each of provides independent power to a circuit. So the search space consists of independent states, which can be checked quicker by grouping. Flip five at a time, and we’ll have toast in no time.
