Conclusion

    The take home points:
  1. Deep nets are a connectionist framework.
  2. They include a vast range of techniques for learning things from large data sets.
  3. The framework is flexible, but often takes a lot of work to learn things for a particular problem.
  4. If you're interested in real neural nets, Chris is happy to talk to you.
  5. If you're interested in having some machine learning folks work with you on particular problems, there are a lot of us at Middlesex who are generally open to collaboration.