Thursday, August 14, 2014

Pictures from Austin .Net Users Group August 11, 2014 with Patrick Abbs.

Patrick Abbs gave a great presentation on "Modern Artificial Intelligence for the Curious Layman" to the Austin .Net Users community. Patrick had an intriguing example of using a Neural Net to classify instruments playing a 'C' note.
Here's a few pics:

My jumbled notes:
  1. Neural Networks are a type of Machine Learning algorithm used for object categorization and recognition.
  2. Neural networks work by creating a hierarchical structure of more complex objects and their components.
  3. Fourier Analysis was used in Patrick's example to simplify complex wave forms into simple sine waves.
  4. The human ear component, the cochlea, decomposes sound into a series of sin waves in an unknown manner.
  5. Neurons hold electrical charge. Axons are limit switches.
  6. Apple's Siri is run by a Hidden Markov algorithm, while Google is investing in Neural nets.
  7. "Equational vs. Structural" was his main theme.
  8. Patrick had three types of Neurons in his example program: 1. Input Nueron 2. Abstration Neuron 3. Monitor Neuron
  9. contacts: twitter @patricks github/patricks

No comments: