

  There are  four parameters  that   can  be set in a
  hopfield network:

	  - Learning Rate
	  - Decay
	  - Treshold
	  - Cycles

  The learning rate should  be  a positive number. It
  determines the  speed with  which the  weights  are
  changed during learning. If set  to a high  number,
  the weights are changed  very fast   and a  pattern
  will  be   learned  quickly,   at  the  cost of the
  patterns that have been learned.  The default value
  is 1.

  The decay should be a positive  number smaller than
  1. The default is 0.5. It determines the speed with
  which  a  network unlearns a pattern.   If set to 1
  the network unlearns the pattern in one cycle.

  The treshold should be number between -1 and 1. The
  default is   0. The  treshold determines how   much
  input  a  unit requires for  it  to  become  active
  (activation 1). When  a network  does not recognize
  the pattern, it may be because the treshold  is set
  too  high. When a network  recognizes more than one
  pattern at one and the same time, the  treshold may
  be set too low.

  The cycle-parameter should  be a  positive  number.
  The  default is  2. If  it set  to a high  number a
  pattern  the weights will  be updated several times
  and the prvious patterns will be forgotten a little
  more each time.
