Artificial Intelligence Night schedule

Unscheduled

• Adding machine learning to a web app

  by Richard Dallaway

  ...an insight into a nuts-and-bolts of introducing machine learning to a web app, looking at the process, what you need to consider and the gotchas.

  Richard has a BA in AI from Sussex, and focussed on machine learning to get his DPhil. He worked for the machine learning group at Logica, coding and consulting on a range of AI problems including financial prediction, basket analysis and fraud detection for corporate clients. He's now the director of Spiral Arm, building funky bespoke web apps.

  • Machine Learning
  Coverage slide deck

• IBM Watson

  by Dale Lane

  • What problem is the Watson / DeepQA project trying to tackle? And why is it a difficult area?
  • What were IBM's previous efforts in this space, and how did they lead to the Watson project?
  • Why was Jeopardy chosen for the project?
  • A crash course in how Watson works - the hardware architecture, the overall software architecture, and peeking under the hood at some of the more interesting algorithms
  • Looking forward at what might come next for Watson.
  • IBM Watson

• Past A.I. projects and "An AppStore for A.I."

  by Ian Ozsvald

  Ian is an A.I. consultant who has enjoyed applying clever algorithms to hard problems in industry for a decade. He lectures sometimes at Sussex Uni and teaches High Performance Python programming at Python conferences.

  He'll talk through some past A.I. projects that will include:

  • discovering who is at a venue using face detection and FourSquare logins
  • an open source face tracking robot head (featured on The Gadget Show)
  • using O.C.R. to augment the OpenPlaques website by reading photos of English Heritage plaques

  Then Ian will present a new project - a sort of AppStore for A.I. and complex algorithms that are designed to make it easy for developers to test ideas and build new services. If things go well, Ian will also run through a demo of the forthcoming web API.