4 - Brain as Information System Flashcards
What was Claude Shannon’s contribution at the Bell Labs (1948) to information theory?
Published mathematical theory of communication.
Measure of redundancy (high value and low value) of information.
What is Bayesian Framework?
Framework for understanding probabilistic systems.
Points:
- brain required to make guesses from uncertain information.
- to solve this uses current inputs and existing knowledge to provide best guess.
Explain the inputs and outputs of the brain as a processing system
External inputs; sensory information
Internal inputs; genetically coded (nature), stored information (nurture)
External outputs; behaviour, modifying stored information (memories)
What is a representation?
Information stored in a particular format
Not all representations are equal (computer info of photo)
What is a process?
An algorithm that changes the form of the representation
What are Marr’s levels of explanation?
Three levels to any problem;
- Computational theory; what are we trying to accomplish?
- Representation and Algorithm; what are the steps we will take to reach this goal? (inputs:ingredients, algorithms: recipe, outputs:goal)
- Hardware implementation; how are we physically going to accomplish this task? Mechanical machine, biological system.
There’s interactions between levels and science (multiple solutions for the same model)
What did Marr’s Levels of Explanation achieve?
Provided general framework on how he believed people should investigate any process system
Very influential because it’s broad.
How are the photoreceptors in the retina distributed ?
Most cones (day vision) in the fovea with rods distributed across the retina.
Explain Simple Circuits in information processing in the retina
Each photoreceptor gets it’s own neuron.
Each neuron indicates a single point of stimulation.
Great acuity
Uses lots of resources because of 1:1 ratio
E.g. Foveal cones
Explain Convergent Circuits
Multiple receptors excite a single neuron.
It’s bad for acuity, but good sensitivity. Uses less resources.
E.g. peripheral neurons in the eye.
What is the effect of simple and convergent circuits on foveal representation?
High acuity and colour vision in central vision, but poor sensitivity.
Good for reading and performing detailed tasks
Can’t see colour at night and can’t see with weak stimuli
What is the effect of simple and convergent circuits on periphery representation?
Poor acuity but good sensitivity
Can see at night
How does the brain perceive objects?
Brain optimises the information it is given.
E.g. faces and cars are easily confused in conceptual space, so brain needs to be able to separate the faces (in IT space)
