exercise2():
  Seek parameters for logistic regression example.
  We have M samples of gold coins.
  Data includes g, the weight in grams, and
  y, a label of 0 for fake coins, 1 for real coins.
  We seek a logistic regression model
    y = 1 / ( 1 + exp ( - ( w(0) + w(1) * g(:) ) )
  for which we need to estimate the weights w.

  Use gradient descent.

  Estimated weights W = (-3.88351,6.8486)
  Cutoff value for normalized data is  0.5670512127712141
  Cutoff value  is  33.53947305405286
  Graphics saved as "exercise2.png"

exercise2():
  Normal end of execution.