Please enable JavaScript.
Coggle requires JavaScript to display documents.
Task 6 M&M Dynamic systems :3 !! (Dynamic System Models --> picks…
Task 6 M&M Dynamic systems :3 !!
Dynamic System Models
--> picks up many things not factored in by CRUm
---> BUT still not enough evidence to replace CRUM cause cant explain as many things like (problem solving, learning , language)
3 ways (from summary)
metaphorical use of dynamic systems
connectionist models described as dynamic systems
small number of variables +quations
mind and world not separate but interact in a complex and dynamic everchanging system :3 !!
-->our perception is deeply influenced by what we perceive and how we react to the world
—> CRUM doesnt take that into account !
Advantages of dynamic systems
Can better explain time
better explains how interaction of mind with world changes behavior
better explains "non-representational" human behavior
--> e.g. moods, sleep, motor control (which influenced by interaction of environment and human mind)
IMportnat !!
dynamic systems not enough support to be an alternative to CRUM cause CRUM can explain
problem solving, learning , language ,
—> dynamic systems havent adressed how these things might work
—> however if wished to also factor in complex human influences (peer pressure, predisposedness, environment etc) dynamic systems or lending form dynamic systems to combine with connectionist model = good approach :)!!
— see green note as that is good example !!
3D cube (CUSP catastrophy :D!!)
--> with the fold on the surface and love and behavior and relationships
attractor
--> is a stable place within a system that the system tends to settle in
--> a system can have multiple attractors, so multiple stable states
Multiple attractors
-> system can have multiple attractors / stable places
state space
--> variables of a system that are used to measure it
--> eg windspeed humidity, temperature in a weather system
phase transitions
--> = move from one stable position/ attractor to another one
#
stable patterns
Intentionality
-->intentionality = property of “being about something” / being deliberate / purposeful !!
--> representation in comuters = independend of the world / interaction with the world and thus dont understand the purpose / intentionallity of it
Situated action
thought is not a solitary, disembodied occurrence, but rather occurs in individuals who interact with a physical world.
chaos
--> system sensitive to small variable input
--> small change in values can lead to HUGE difference in outcome
Bifurcation point :D !!
--> point at which catastrophic change happens if entered form one side and exiting it on the other side :3 !! and vice versa :3 !!
, ,
splitting factor
--> factor which at moderate normal factor levels leads to divergence/ change in behavior
normal factor
--> factor which increases monotonically with behavior while the splitting factor remains constant
hysteresis :D!!
--> point at which switch in behavior takes longer when apprached from one side of the normal factor (e.g 100% love towards 0% love) than from the other side (dependence on history)
having a position in the beginning makes you stick longer with that position as if you had a different starting point
read new synthessi cause its the line picutre in the course manual :3 !!
add the figure descripton lol
computational approach
Challenges for Cognitive Science
CRUM
-->CRUM can explain as many things like (problem solving, learning , language)
CRUM limitations:
---> 4 ways to deal with these:
--> CRUM = Computational representational understanding of mind (CRUM)
--_> CRUM all these mistakes + What we can do to fix !!
--> lets take step 3 and add non Crum elements to Crum e.g. Dynamic systems !!
-------> chapter 12 lets take connectionism and add dynamic systems to make them super duper goood :3 !!
1. brain challenge
--> CRUm ignores facts how brain actually thinks
2. emotion challenge
--> doesnt factor in emotions
3. consciousness challenge
--> doesnt factor in consciousness
4. body challenge
--> doesnt factor in body / physiological changes
5. world challenge
--> doesnt factor in physical environment / context
6. dynamic systems challenge
--> doesnt factor in that mind = dynamic system
7. social challenge
--> doesnt factor in social influences
Answer to limitations:
Deny these challenges
Expand CRUM
-->by adding new computational + representational ideas to meet the challanges
Supplement Crum
--> by adding noncomputational, nonrepresentational con- siderations that together with CRUM can meet the challenge
4. Abandon Crum
#
Expanding CRUM
--> adding all the missing abilities to CRUm to be able to perceive what humans perceive to gain new more accurate representations!
--> eg. by adding ability to smell, see, hear and touch, + feeling of arm when swinging hammer etc etc all needed
Supplementing CRUM
--> giving additional representing input eg. how visual inout is perceived and processed in human brain
---> could lead to better functioning through understanding like human does more accurately
Abandoning CRUM
--> not good idea becasue it can already explain and do many things, planning, learning and language !
#
#
Comparative Evaluation
Neurological Plausibility
Psychological Plausibility
--> different experimental results fit best with different representational theories
--> there is no best fit its all task dependent!
rule based to cognitive task
connectionist model to cognitive task involving incremental learning +parallel constraint
Computational Power
Representational Power
--> 3 concepts of blabla blaaaa
Practical Applicability
6 approaches to representation and computation
FOCUS ONLY ON IF THEN RULES AND CONNECTIONISM IN CHAPTER 8 and look at all benefits so wehre eah si good and where each is worse than the other !!