[Colloquium] Talks TODAY at TTI-C, International House: 5/23-5/27

[Katherine Cumming]

(3)  TTI-C Show and Tell Series
Speaker: Derek Dreyer, TTI-C
Speaker's Homepage:  http://www.tti-c.org//dreyer.html
Time:  Tuesday, May 24 @ 12:15pm - Lunch Provided by TTI-C
Location:  TTI-C Conference Room 
Title: Recursive Abstract Types
Abstract:
Existential types provide a simple and elegant foundation for
understanding generative abstract data types, of the kind supported by
the Standard ML module system. However, in attempting to extend ML with
support for recursive modules, we have found that the traditional
existential account of type abstraction does not work well in the
presence of mutually recursive module definitions. The key problem is
that, in recursive modules, one may wish to define an abstract type in a
context where a name for the type already exists. Existential types
provide no way for the implementor of an existential to connect her
hidden type representation to a pre-existing abstract type name.

We propose a novel account of type abstraction that resolves this
problem. The basic idea is to separate the act of generating a name for
an abstract type from the act of defining its underlying representation.
To define several abstract types recursively, one may first
``forward-declare'' them by generating their names, and then define each
one secretly within its own defining expression. In contrast to the
usual continuation-passing interpretation of existential types in terms
of universal types, our account of type abstraction suggests a
``destination-passing'' interpretation. Briefly, instead of viewing a
value of existential type as something that creates a new abstract type
every time it is unpacked, we view it as a function that takes as input
a pre-existing undefined abstract type and defines it. By leaving the
creation of the abstract type name up to the client of the existential,
our approach makes it significantly easier to link abstract data types
together recursively.

In this talk, I will not assume that you know anything about existential
types or recursive modules, and will introduce the relevant ideas from
first principles as necessary. 
 
(4) Learning Theory Program/ML Summer School 
 
Speaker:  Mark Liberman, University of Pennsylvania
Speaker's Homepage: http://www.ling.upenn.edu/~myl/
Time:  Tuesday, May 24 @ 2:00pm
Location:  International House-Assembly Hall
Title:  Categorical Perception + Linear Learning = Shared Culture
Abstract:
In a group of entities who learn by observing one another's behavior,
some simple assumptions about the nature of perception, the nature of
individual beliefs, and the nature of learning lead naturally to
collective convergence on a random set of shared beliefs, without any
structure of authority or any explicit collective decision process. This
process will be exemplified in the case of a simple model for developing
word pronunciations. 
 
(5) Learning Theory Program/ML Summer School
Speaker:  Fei Sha,  University of PA
Speaker's Homepage:   <http://www.cis.upenn.edu/~feisha/>
http://www.cis.upenn.edu/~feisha/
Time:  Tuesday, May 24th @ 2:00pm
Location:  TTI-C Conference Room
Title:  Efficient Parameter Estimation for Large-Scale? Conditional
Random Fields
Abstract:
Conditional random fields for sequence labeling offer advantages over
both generative models like HMMs and classifiers applied at each
sequence position.<BR.
Among sequence labeling tasks in language processing, shallow parsing
has received much attention, with the development of standard evaluation
datasets and extensive comparison among methods.

We show here how to train a large-scale conditional random field to
achieve performance as good as any reported base noun-phrase chunking
method on the CoNLL task, and better than any reported single model.
Improved training methods based on modern optimization algorithms were
critical in achieving these results.

Space is limited. For more information on this session for Machine
Learning Summer School Participants Click
<http://chicago05.mlss.cc/tiki/tiki-read_article.php?articleId=10>
Here.
 
(6) Learning Theory Program/ML Summer School 
Speaker:  Antonio Galves, University of Sao Paolo
Speaker's Homepage: http://www.ime.usp.br/~galves/hp/
Time:  Tuesday, May 24th @ 3:10pm
Location:  International House-Assembly Hall
Title:  Fingerprints of Rhthm in Natural Language
Abstract:
This talk reviews a list of recent results on the rhythmic classes
hypothesis produced by the Tycho Brahe Project. I start with results on
the rhythmic classification of speech data based on the speech sonority.
Then, I address the question of the identification of fingerprints of
rhythm in Brazilian and European written texts. I conclude discussing
the role fingerprints of rhythm may play in language acquisition and
change.
 
If you have questions, or would like to meet the speaker, please contact
Katherine at 773-834-1994 or kcumming at tti-c.org.   For information on
future TTI-C talks and events, please go to the TTI-C Events page:
http://www.tti-c.org/events.html.  TTI-C (1427 East 60th Street,
Chicago, IL  60637)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
------------------------------------------------------------------------
-----
If you have questions, or would like to meet the speaker, please contact
Katherine at 773-834-1994 or kcumming at tti-c.org. For information on
future TTI-C talks or events, please go to the TTI-C Events
<http://ttic.uchicago.edu/events/events_dyn.php>  page. TTI-C:  1427
East 60th Street (University Press Building 2F), Chicago, IL  60637
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20050524/f102b429/attachment.htm