Conclusion for Deep Nets

• Take Home Points
  1. Deep Nets involve many layers of neurons.
  2. There are a range of them, but Hinton's Deep Belief Net is an important early example.
  3. They work on large data sets.
  4. There are many algorithms used and many different models of units.
  5. Using these, much less developing them, is very complex.
• Reading: For this week read the Deep Learning Wiki
• Reading: for next week is the Knowledge rperesentation in the semantic web for Earth and environmental terminology (SWEET) by Raskin and Pan (2005), and XML Tutorial.