exercise3:
  Seek parameters for logistic regression example.
  We have N examples of D-dimensional data X.
  For a model
    y(x) = 1 / ( 1 + exp ( - ( w(0) + w(1)*h + w(2)*t ) ) )
  determine the weights w.

  Estimated weights W = (25.2394,-0.160188,-0.217257)
  Graphics saved as "exercise3_data.png"
  Graphics saved as "exercise3_fit.png"

exercise3():
  Normal end of execution.