PSY2002 W4 Models (L) Flashcards
What is a model?
simplified (or idealized) representation of a more complex thing
What is the Marr’s three levels of analysis?
- Computation – the problem being solved
- Algorithm – the steps/rules to solve it
- Implementation – the actual machinery
What is abstraction?
generating general rules and concepts from specific information
What is simplification ?
making something simpler
What is algorithm (rules)?
Rules: What representations and algorithms can we generate, given specific neural circuits?
What is top-down appraoch?
Problem: What is the problem we’re trying to solve
Rules: What representations and algorithms can solve this problem?
Implementation: How can these representations and algorithms be implemented in neural circuits?
What is bottom-up appraoch?
Implementation: the machinery of neural circuits
Rules: What representations and algorithms can we generate, given specific neural circuits?
Problem: What problems are solved by these algorithms?
Marr (1982): “…an algorithm is likely to be understood more readily by understanding the nature of the problem being solved than by examining the mechanism (and the hardware) in which it is embodied.“
What are Statistical models?
a mathematical relationship between variables, that hold under specific assumptions
What are Theoritical models?
(in cognition) a description of the relationship between different mental processes, that makes assumptions about the nature of these processes
What is behaviourism?
Input > Mind [black box (you can’t understand] > Output (behaviour)
Behaviourists want to understand what is going on in the mind by changing the inputs and observing the outputs.
What is Cognitive “box-and-arrow” models?
Box-and-arrow can be called informal cognitive models.
Models that describe the relationship between different mental processes, under the assumption that the mind operates like multi-staged information-processing machines. Box and arrow models started off simple, but can gradually become quite complex.
What are Formal Cognitive Models?
mathematical description of the relationship between mental processes. Usually expressed through computer code
Models are characteristed by what?
All models simplify and make things more abstract.
Simplification and Abstraction
When we create a model, we acknowledge that we’re not going to describe all the information we’re describing, only the parts we think are critical for what we’re trying to represent.
What is the right level of abstraction of a model?
depends on the question we are asking and/or what we’re trying to convey.
“All models are wrong, but some are useful. “- George Box (1976)
“The map is not the territory. “- Alfred Korzybski (1931)
What do models in science must produce?
Predictions.
These predictions can be directional or numerical. Models that provides numerical predictions can be more or less accurate.
Karl Popper - Non-scientific theories explain after the fact but cannot provide falsifiable predictions
What are specific numerical prediction?
They do not give explanations but it steams from a theory (explanation) - theoretical model
Difference between statistic model and theoretical model (trying to explain, with a theory behind it)
How do we use models to predict and explain?
Framework <> theory <> model <> hypothesis <> data
What is framework?
Models to predict and explain
The conceptual system that defines terms and provide context
Cognitive psychology
What is a theory?
Models to predict and explain
A scientific proposition that provides relations between phenomena
Early-Selection theory
A schematic representation of a theory, more limited in its scope
Broadbent’s Filter model
What is a hypothesis?
Models to predict and explain
A narrow testable statement
Irrelevant stimuli that contain target-defining feature will be automatically detected
What is data?
Models to predict and explain
Collected observation, often as part of an experiment
New “gorilla’ experiment, detection rates, t-tests.
What is model ?
Model to explain and predict
A schematic representation of a theory, more limited in its scope
Broadbent’s Filter model
What are explanation without exact prediction?
Models of schizophrenia can indicate causes but cannot (yet?) predict individual cases.
The model be able to predict group differences, but not individual cases.
What is prediction without explanation?
Some models can predict whether an individual will develop Alzheimer, even though we still aren’t close to understand the factors that explain Alzheimer
What are informal cognitive models?
a verbal description of the relationship between different cognitive procedure:
Where often some assumptions are implicit.
Often provides only directional predictions
What are formal/computational models?
Formal/computational models: a mathematical description of the relationship between different cognitive procedure, often instantiated via computer program/simulation:
Assumptions are explicit.
Often provides numerical predictions
How do formal models explain?
Specification – a formal description of the relations described by a theory - The formal model, comprised of symbolic representations
Implementation – a specific instantiation of a specification – a computer program ale to simulate and predict numerical outputs from input.
Why are formal models more specific in prediction?
Gottfried Wihelm Leibniz, 1685
“Let us calculate, without further ado, and see who is right”
By having a numerical simulation, we can see if the model provides unreasonable predictions (easier to reject bad models). Can help us select which experiments to perform.
By having numerical predictions, we can provide a more subtle form of hypothesis testing. We can see how close a model is to predicting an actual result.
Why do formal models have more counter-intuitive prediction?
A model can more clearly describe which predictions follow from a model.
With informal models, it’s hard to notice when they make counter-intuitive predictions. Formal models clearly produce such predictions.
Why do formal models benefit of explicit assumptions?
By making assumptions explicit, we can reveal unanswered questions, flaws in our reasoning, contradictory or unreasonable assumptions. Feynman’s blackboard (“what I cannot create, I do not understand”)
How does David Marr’s Levels of analysis understand the brain?
Problem: We can only ever hope to sample from a tiny fraction of its activity, in a tiny fraction of a bit of brain
The way to make sense of brain data was to break any brain problem into three levels:
- Computation – the problem being solved
- Algorithm – the steps/rules to solve it
- Implementation – the actual machinery
