Jonathan Burn (OC) works in merchant bank IT and marketing. He is, in addition, a funk musician and mathematician, was a primary school teacher, and did a degree in Economics at the LSE in his own time for fun. He also has a masters in Artificial Intelligence from Imperial.
With an engaging and clear set of slides, Jon took us carefully through what sound is, how computers store the information contained in sounds, and how they can be set up to learn the best ways of identifying new sounds. This is done mainly through the ‘random forest’ technique: a way of generating questions to ask about a new sound which most accurately and efficiently put it into the right group (e.g. robin or blackbird).