Introduction to Cognitive Modeling

almondpitterpatterΤεχνίτη Νοημοσύνη και Ρομποτική

23 Φεβ 2014 (πριν από 3 χρόνια και 6 μήνες)

77 εμφανίσεις

Fall 2004 Cognitive Science 207



Introduction to

Cognitive Modeling

Praveen Paritosh

Overview


Who we are


Course mechanics


What is cognitive modeling?


Syllabus


Homework Zero


Who we are


Praveen

Paritosh


Brian

Kyckelhahn


Kate

Lockwood

Mechanics


Combination of lectures and discussions


Weekly homeworks


Midterm will be Thu October 21
st
, in class


Final exam will be Fri December 10, 12pm
-
2pm


Communications


Class web site =

http://www.cogsci.northwestern.edu/courses/cg207/


To contact Brian, Kate or Praveen re class
matters:

cogsci207
-
staff@cs.northwestern.edu


For class discussions, we will use the
discussion forums in Blackboard

https://courses.northwestern.edu/webapps/login


Grading


Midterm: 20%


Final exam: 30%


Reading/Modeling Assignments: 50%

Reading papers


No textbook, but a collection of research
papers.


We want you to READ the papers.

Critiques


For each paper, three one sentence long
critiques


of what is wrong with the paper.


Due at the beginning of the Tue
(Discussion) class.


Will be used as a basis for the discussion, so
be prepared to defend your critique!


Will account for a third of your grade.


Classes


Thursday:


Lecture


Readings assigned


Tuesday:


Critiques due before class


Discussion based on critiques and readings


Modeling homework assigned, due following Tuesday.


Modeling Assignments


Turned in via email to

cogsci207
-
staff@cs.northwestern.edu


No hardcopies or email to other addresses


ASCII or HTML preferred, followed by PDF or
Word. (If HTML, must be self
-
contained:
Broken links will lose you points)


Late homeworks will be downgraded


All work you turn in must be your own.

Reading assignments due beginning of discussion class
on Tuesdays. Bring hardcopy of critiques to class.



What is mind?


One of the deepest questions humanity has
asked


Many fields have tried to answer it


Philosophy


Psychology


Linguistics


Biology (evolutionary, neuroscience, …)





It’s probably a computation


A key insight


Productive, since it raises many questions


What’s a computation?


What kind of computation?


Operating over what kinds of data?


On what sort of system is it being carried out?

Artificial Intelligence


Goal: To understand the nature of
intelligence


In whatever kind of system can exhibit it,
including people


Early successes inspired (and inspired by)
comparison with human cognition


Solving problems, playing chess, parsing
sentences, seeing in simple scenes, …

Cognitive Science


Born out of the computational insight


Computation could provide a new theoretical language
for cross
-
discipline communication


Meeting ground for fields traditionally concerned
with studying cognition



Multidisciplinary

field


Each field has theoretical constructs to share


Each field has its own empirical methods for testing
ideas


Deeper insights come out of their interactions

Can a machine think?


What you will learn


A basic understanding of how computation
can be used to model phenomena in
cognitive science


Crucial for all cognitive scientists, since
computation is the theoretical language of the
field


Facilitate working with computational
modelers, if you aren’t going to become one


Good start to becoming a computational
modeler, if that’s what you want to do.

Methodology


What does it mean to model thinking?


Turing test and its limitations


Chatterbots

Knowledge representation


How can computers know things?


Overview of how reasoning systems work


An introduction to predicate calculus


A high
-
level tour of the Cyc knowledge base


Ontology


Microtheories


Naïve physics


How can we model our everyday
understanding of the physical world?


Qualitative representations as formalization of
conceptual knowledge


Vmodel software



Natural language processing


How can we model the understanding of
language?


Guest lecturer: Chris Riesbeck


Music Cognition


Representations of how we understand/
interpret music.


Guest Lecturer: Bryan Pardo

Analogy and similarity


How do we reason and learn from analogies
and metaphors?


Gentner’s structure
-
mapping theory


Computational simulations of it


Learning and education


How do we learn new theories and skills?
Can we use these models to teach?


Production
-
rule models of skill


CMU work on intelligent tutoring systems


Emotions and Consciousness


How can we study them as scientists?


Norman
et al
’s model of emotions in cognitive
architecture


McDermott’s analysis of consciousness

Homework Zero


Due Tue, Sep 28, noon. Email to

cogsci207
-
staff@cs.northwestern.edu

as always.


Questions:

1.
Why are you taking this course?

2.
What cognitive phenomena would you most like
to model?

3.
Have you had any background in programming
or computing more generally?


Task:


Post a comment to one of the Discussion Boards for
the course in Blackboard


Readings


Turing, A. M. "
Computing Machinery and
Intelligence
," Mind, New Series, Vol. 59, No. 236.
(Oct., 1950), pp. 433
-
460. (also available
here
).





Minsky, M.
"Why people think computers
can't"
.


AI Magazine
, Fall, 1982.



Miller, G. "
The Cognitive revolution: A historical
perspective
", Trends in Cognitive Sciences, 7(3),
March 2003.