COGS 200 Midterm Questions With Complete Solutions
Prototype
Each category (e.g. "bird") has this- a mental image of what the
most typical object in that category looks like.
exemplar model
If an object resembles this, we add it to the category, and it
becomes one itself
exemplar
Instead of one prototype, we have many of these. Every object
we have ever seen within a category, becomes this for that
category
Nosofsky's task
Categorize cartoon faces, which differ based on a small number
of measurable variables. Two categories, A and B, based on a
mathematical rule - but only the person running the experiment
knows the rule. Computers (or humans) see faces and try to
guess which category they belong to.
Most humanlike model
According to Nosofsky's task, the exemplar model since it has a
similar learning rate to a human and the model make the same
kinds of mistakes as a human
Cognitive architecture
,Attempts to model the whole mind, or a large section of the
mind, by connecting different processes together
Cognitive architecture use
See how mental processes interact with each other for tasks like
learning, memory, and decision making
Parts of a Cognitive Architecture
Short- and long-term memories that store agents' beliefs, goals,
and knowledge, performance mechanisms that use the contents
of memory for some tasks, such as problem solving, and
learning mechanisms that can alter or add to the contents of
memory
Evaluating Cognitive Architectures
Generality and versatility, rationality and optimality, efficiency
and scalability, reactivity and persistence, improvability, and
autonomy and extended operation
Generality and Versatility
Can the CA perform many different thinking tasks? Can it learn
to perform a new task?
Rationality and Optimality
Does the CA do its task well?
Efficiency and Scalability
Does the CA take up an unreasonable amount of processing
power or time?
Reactivity and Persistence
, How does the CA respond to unexpected situations
Improvability
Will it be easy to update the CA in response to new theories or
experimental data bout cognition? Is it easy for the CA to update
its own knowledge and skills?
Autonomy and Extended Operation
Can the CA decide for itself what to do or work? How long can
it keep doing this?
Recursion
A statement that refers to itself or a function that "calls" another
version of itself or a sentence in which the same predicate
appears in the head and in the body. It is a powerful tool in
computer programming
Recursion example
ancestor(X,Y) :- X is directly above Y, in the family tree, by any
number of generations
Base case
The condition under which a recursive function returns without
calling itself, thereby ending the recursive calls.
Recursive case
The condition under which a recursive function returns by
calling itself, thereby continuing the recursive calls
Base case example
ancestor(X,Y) :- parent(X,Y)
Prototype
Each category (e.g. "bird") has this- a mental image of what the
most typical object in that category looks like.
exemplar model
If an object resembles this, we add it to the category, and it
becomes one itself
exemplar
Instead of one prototype, we have many of these. Every object
we have ever seen within a category, becomes this for that
category
Nosofsky's task
Categorize cartoon faces, which differ based on a small number
of measurable variables. Two categories, A and B, based on a
mathematical rule - but only the person running the experiment
knows the rule. Computers (or humans) see faces and try to
guess which category they belong to.
Most humanlike model
According to Nosofsky's task, the exemplar model since it has a
similar learning rate to a human and the model make the same
kinds of mistakes as a human
Cognitive architecture
,Attempts to model the whole mind, or a large section of the
mind, by connecting different processes together
Cognitive architecture use
See how mental processes interact with each other for tasks like
learning, memory, and decision making
Parts of a Cognitive Architecture
Short- and long-term memories that store agents' beliefs, goals,
and knowledge, performance mechanisms that use the contents
of memory for some tasks, such as problem solving, and
learning mechanisms that can alter or add to the contents of
memory
Evaluating Cognitive Architectures
Generality and versatility, rationality and optimality, efficiency
and scalability, reactivity and persistence, improvability, and
autonomy and extended operation
Generality and Versatility
Can the CA perform many different thinking tasks? Can it learn
to perform a new task?
Rationality and Optimality
Does the CA do its task well?
Efficiency and Scalability
Does the CA take up an unreasonable amount of processing
power or time?
Reactivity and Persistence
, How does the CA respond to unexpected situations
Improvability
Will it be easy to update the CA in response to new theories or
experimental data bout cognition? Is it easy for the CA to update
its own knowledge and skills?
Autonomy and Extended Operation
Can the CA decide for itself what to do or work? How long can
it keep doing this?
Recursion
A statement that refers to itself or a function that "calls" another
version of itself or a sentence in which the same predicate
appears in the head and in the body. It is a powerful tool in
computer programming
Recursion example
ancestor(X,Y) :- X is directly above Y, in the family tree, by any
number of generations
Base case
The condition under which a recursive function returns without
calling itself, thereby ending the recursive calls.
Recursive case
The condition under which a recursive function returns by
calling itself, thereby continuing the recursive calls
Base case example
ancestor(X,Y) :- parent(X,Y)
