It crops up all over the place: you have some mathematical function, and you would like to know for what inputs do you get the “best” output. For example, when designing an aeroplane, you might be able to change the size and shape of the wings, and you’d like to know how much lift they would generate. Similar problems appear in fields as diverse as linguistics and physics, and optimisation is at the core of machine learning.

In this session we look at what we mean by optimising a function, what maths lies behind it and some common methods for function optimisation.

You can find the slides for this talk here.