Model Representation ll

  |   Source
Model Representation ll
  • how to do it efficiently (vectorized)
  • why neural networks works,  how is it a good idea

  • rename the content of g as z
  • g is sigmoid function of multiplication wise element of z
  • remember that a1 is units in layer 1
  • a3 is  the final unit for hypothesis
  • the computation of formula written on the right side is the simple way if calculating h of x.