[1]Principia Cybernetica Web

                              Learning and Model-Building

   Cybernetic epistemology is in essence [2]constructivist: knowledge cannot be
   passively absorbed from the environment, it must be actively constructed by the
   system itself. The environment does not instruct, or "in-form", the system, it
   merely weeds out models that are inadequate, by killing or punishing the system
   that uses them. At the most basic level, model-building takes place by
   variation-and-selection or [3]trial-and-error. Let us illustrate this by
   considering a primitive aquatic organism whose [4]control structure is a slightly
   more sophisticated version of the thermostat. To survive, this organism must
   remain in the right temperature zone, by moving up to warmer water layers or down
   to colder ones when needed. Its perception is a single temperature variable with
   3 states X = {too hot, too cold, just right}. Its variety of action consists of
   the 3 states Y = {go up, go down, do nothing}. The organism's control
   [5]knowledge consists of a set of perception-action pairs, or a function f: X
   ;->Y. There are 3^3 = 27 possible such functions, but the only optimal one
   consists of the rules:
     * too hot ;-> go down
     * too cold -> go up
     * just right -> do nothing.

   The last rule could possibly be replaced by either just right -> go up or just
   right -> go down. This would result in a little more expenditure of energy, but
   in combination with the previous rules would still keep the organism in a
   negative [6]feedback loop around the ideal temperature. All 24 other possible
   combinations of rules would disrupt this stabilizing feedback, resulting in a
   runaway behavior that will eventually kill the organism.

   Imagine that different possible rules are coded in the organism's genes, and that
   these genes evolve through random mutations each time the organism produces
   offspring. Every mutation that generates one of the 24 combinations with positive
   feedback will be eliminated by natural selection. The three negative feedback
   combinations will initially all remain, but because of competition, the most
   energy efficient combination will eventually take over. Thus internal
   [7]variation of the control rules, together with natural [8]selection by the
   environment eventually results in a workable model.

   Note that the environment did not instruct the organism how to build the model:
   the organism had to find out for itself. This may still appear simple in our
   model with 27 possible architectures, but it suffices to observe that for more
   complex organisms there are typically millions of possible perceptions and
   thousands of possible actions to conclude that the space of possible models or
   control architectures is absolutely astronomical. The information received from
   the environment, specifying that a particular action or prediction is either
   successful or not, is far too limited to select the right model out of all these
   potential models. Therefore, the burden of developing an adequate model is
   largely on the system itself, which will need to rely on various internal
   heuristics, combinations of pre-existing components, and subjective [9]selection
   criteria to efficiently construct models that are likely to work.

   Natural selection of organisms is obviously a quite wasteful method to develop
   knowledge, although it is responsible for most knowledge that living systems have
   evolved in their genes. Higher organisms have developed a more efficient way to
   construct models: learning. In learning, different rules compete with each other
   within the same organism's control structure. Depending on their success in
   predicting or controlling disturbances, rules are differentially rewarded or
   reinforced. The ones that receive most reinforcement eventually come to dominate
   the less successful ones. This can be seen as an application of control at the
   metalevel, or a [10]metasystem transition, where now the [11]goal is to minimize
   the perceived difference between prediction and observation, and the actions
   consist in varying the components of the model.

   Different formalisms have been proposed to model this learning process, beginning
   with [12]Ashby's homeostat, which for a given disturbance searched not a space of
   possible actions, but a space of possible sets of disturbance -> action rules.
   More recent methods include neural networks and genetic algorithms. In genetic
   algorithms, rules vary randomly and discontinuously, through operators such as
   mutation and recombination. In neural networks, rules are represented by
   continuously varying connections between nodes corresponding to sensors,
   effectors and intermediate cognitive structures. Although such models of learning
   and adaptation originated in cybernetics, they have now grown into independent
   specialisms, using labels such as "machine learning" and "knowledge discovery".

   Reference:
   Heylighen F. & Joslyn C. (2001): "[13]Cybernetics and Second Order Cybernetics",
   in: R.A. Meyers (ed.), Encyclopedia of Physical Science & Technology , Vol. 4
   (3rd ed.), (Academic Press, New York), p. 155-170
     ____________________________________________________________________________

   [14]Copyright© 2001 Principia Cybernetica - [15]Referencing this page

   Author
   F. [16]Heylighen, & C. [17]Joslyn,

   Date
   Sep 3, 2001

                                       [18]Home
                                       [up.gif]
                           [19]Metasystem Transition Theory
                                       [up.gif]
                                   [20]Epistemology

                                          Up
                           [21]Prev. [4arrows.gif] [22]Next
                                         Down
     ____________________________________________________________________________
   ____________________________________________________________________________

                                    [23]Discussion
     ____________________________________________________________________________

                                  [24]Add comment...

                                      [space.gif]

References

   1. LYNXIMGMAP:http://pespmc1.vub.ac.be/LEARNMOD.html#PCP-header
   2. http://pespmc1.vub.ac.be/CONSTRUC.html
   3. http://pespmc1.vub.ac.be/TRIALERR.html
   4. http://pespmc1.vub.ac.be/CONTROL.html
   5. http://pespmc1.vub.ac.be/KNOW.html
   6. http://pespmc1.vub.ac.be/FEEDBACK.html
   7. http://pespmc1.vub.ac.be/VARIATIN.html
   8. http://pespmc1.vub.ac.be/SELECT.html
   9. http://pespmc1.vub.ac.be/KNOWSELC.html
  10. http://pespmc1.vub.ac.be/MST.html
  11. http://pespmc1.vub.ac.be/GOAL.html
  12. http://pespmc1.vub.ac.be/CSTHINK.html#Ashby
  13. http://pespmc1.vub.ac.be/Papers/Cybernetics-EPST.pdf
  14. http://pespmc1.vub.ac.be/COPYR.html
  15. http://pespmc1.vub.ac.be/REFERPCP.html
  16. http://pespmc1.vub.ac.be/HEYL.html
  17. http://pespmc1.vub.ac.be/JOSLYN.html
  18. http://pespmc1.vub.ac.be/DEFAULT.html
  19. http://pespmc1.vub.ac.be/MSTT.html
  20. http://pespmc1.vub.ac.be/EPISTEM.html
  21. http://pespmc1.vub.ac.be/EPISTEM.html
  22. http://pespmc1.vub.ac.be/KNOW.html
  23. http://pespmc1.vub.ac.be/MAKANNOT.html
  24. http://pespmc1.vub.ac.be/hypercard.acgi$annotform?

[USEMAP]
http://pespmc1.vub.ac.be/LEARNMOD.html#PCP-header
   1. http://pespmc1.vub.ac.be/DEFAULT.html
   2. http://pespmc1.vub.ac.be/HOWWEB.html
   3. http://pcp.lanl.gov/LEARNMOD.html
   4. http://pespmc1.vub.ac.be/LEARNMOD.html
   5. http://pespmc1.vub.ac.be/SERVER.html
   6. http://pespmc1.vub.ac.be/hypercard.acgi$randomlink?searchstring=.html
   7. http://pespmc1.vub.ac.be/RECENT.html
   8. http://pespmc1.vub.ac.be/TOC.html#LEARNMOD
   9. http://pespmc1.vub.ac.be/SEARCH.html