1. Matching Problems to AI Solutions
In this video David summarises the different AI techniques we have covered in the course, and discusses how you know when to deploy one of them rather than another.
The problem of selecting an approach for a given problem is well known one, prompting some organisations to create cheat sheets or decision diagrams that can help you decide.
We have covered many of these in this course, but in the video David does mention a few that we have not looked at, if you are interested you may want to investigate some of these in your own time.
Its also worth knowing that sometimes, even within a given dataset, different machine learners will have different performance depending on the context. For this reason, AI developers sometimes use ensemble learning, which is where you train multiple machine learners and then use them together to get an answer (for example, by allowing them to vote).