Plasticity and evolution
Last week, we looked at...
Main questions for today
This week we will look at, ...
In simple animals, the `brain' is determined completely by the DNA
In larger animals other mechanisms of brain development must operate
Self-organization in interaction with the environment
Example: Stereo vision
Binocular correspondence is used to calculate depth
Correspondence is calculated by matching contours
A neural network can find the optimal match
Contour lines are matched with a Hebbian learning rule
Such learning must be modulated by global and local factors
The brain must remain plastic and adapt to the changing body during a lifetime
How does the brain develop?
Brain development
Primates are born with all neurons they will ever have
Pseudopodia lead the migrating neuron to its final position along glial fibers
Massive synapse death
Density of synapses in somatosensory, primary motor, prefrontal, and primary visual cortics in monkey
Synapse death is a competitive process
Plasticity can be demonstrated in sensory maps in primates and humans
Overview of sensory cortices
Sensory Homunculus
Reorganization of sensory maps in primate cortex
Ramachandran's procedure
Amputated hand can be sensed on cheek
Sensory maps
Genetic algorithms: artificial evolution
Example of genetic coding
Populations of strings
Fitness
Example of a procedure to determine fitness: camouflage
Major genetic operators
Cross-over
Sources of randomness in genetic algorithms
Genetic algorithms implement a search process
Genetic building blocks
Short schemata are better than long ones and are called `building blocks'
Re-run: The virtual creatures of Karl Sims
Karl Sims' Virtual creatures
Large brains and evolution
Summary
Summary (continued)
Fear is an excellent adaptation to the environment
Email: jaap@murre.com
Home Page: http://www.neuromod.org/courses/public.html
Other information: neuroMod: Home of the Neural and Cognitive Modeling Group at the University of Amsterdam.
Download presentation source