Modelling
and Statistics Learning and Discussion Group (MOSTA)
|
|
|
Papers |
Contribution |
|
|
This peer-driven Modelling and Statistics Learning and Discussion Group at the Geography Department at the University of Calgary is formed with the intent to meet at regular intervals to informally discuss and learn (from papers and from each other's experiences) general modelling and statistical techniques and approaches applicable to the wide array of sub-disciplines in the field of geography. The mission of this group is to provide a forum for the discussion of modelling, statistical, and research design issues, questions and problems in geographical research, to foster colleagiality and an awareness of research activity within the department, and to provide an archive of approaches, solutions, and “best practices” to cooperatively help guide current research projects, or to identify new areas of inquiry.
The
first initiating meeting of this Group was held on June 20th 2006,
with regular bi-monthly meetings during the academic year 2006/2007 (please see discussion schedule for the
previousast year 2006/07). As a result of the
initiating
meeting, an on-line blackboard discussion forum was
created
for people to communicate on any modelling and statistics topic or
question
they might have and to post any materials (papers or presentations) and
questions to be discussed in the fall. Every department member has
access
to this blackboard course called MOSTA_2006, and questions, papers and
ideas can be posted on the discussion forum of this blackboard
site.
If you are interested in joining this on-line discussion group, please
contact Julia Linke (e-mail: jlinke@ucalgary.ca) or Typhenn
Brichierie-Colombi (e-mail: tbrichie@ucalgary.ca).
Meeting
Location:
Earth Science Room 332
Meeting
Time:
Bi-monthly meetings, starting Thursday, September 27, 2007, 11 am to
noon
Meeting
Format:
I)
First 20 minutes:
a) Open forum for reviewing discussions from previous meetings, or
introducing
new topics.
b) Review of the list of topics and questions posted on the
discussion
forum of the blackboard
MOSTA site and follow up of how these issues were solved
II)
Remaining 40 minutes:
“Viewpoint Discussion” (with the posted topic and/or papers
representing
the “viewpoint”), where
MOSTA participants respond to the topic or paper through the sharing of
responses and
experiences related to the topic, or by posing questions.
Facilitator/Moderator
Role:
Anyone can act as a facilitator/moderator, and hence, play a leadership
role in:
a) summarizing Blackboard forums discussions
b) following up on any open forum problems or questions
c) guiding discussion
d) finding/confirming facilitators for subsequent sessions
e) posting notes from meetings on the Blackboard Web forum (e.g., in
the
form of digital
photographs of the white board notes transcribed during meetings)
It is hoped that there will be a diversity of facilitators/moderators
so
that everyone will have a chance to
direct the discussion of topics of interest to them
Some
sample topics to start the group could include:
-
Research Design
-
Sampling Design
- Model selection approaches (p-values versus AIC etc.)
-
Model selection techniques (stepwise, all possible combination,
apriori hypotheses)
-
Bayesian model selection techniques
-
Data dredging/mining
-
Accuracy Assessment of statistical models (ROC and others?)
-
Neural Networks and other black box approaches
-
Autocorrelation in spatial data sets (e.g. wildife data and others)
-
Spatial Smoothing and Interpolation techniques
-
Spatial regression
-
Quantitative versus Qualitative approaches
-
Time series
-
Geocomputation (Cellular Automation, Multi-agent Models)
-
Scripting techniques for statistical programming and modelling
Potentially
as a vision for later, we could also hope to have speakers from the
department
or other departments on certain topics you are interested in or even
develop
ideas for a platform to present some parts of our own work to
each
other.
Peer
Contribution:
New
discussion topics are expected to evolve during group meetings and
should
originate with the research questions facing MOSTA participants. Please
note that many spots are still available for this year, and feel free
to
make your topic of interest known to the group! Also, please see
the current schedule, under the
proviso that it is tentative and hence
flexible for accommodating urgent topics.
This
group will benefit from your involvement and your contribution of
questions
and papers of interest. Please contact Julia Linke (e-mail: jlinke@ucalgary.ca)
or the discussion board on the MOSTA_2006 blackboard course with the
modelling
and statistics topics of your interest or forward any interesting
references
of new and old techniques of innovative or of questionable approaches
that
we can discuss in the group.
Discussion
Schedule: Please follow this
link to the
schedule for the academic year 2007/2008 and then download
the pdf of the paper to be
discussed
for each meeting - note that this schedule can change depending
on
urgency of certain topics and/or on contributions by group members; so
please check this schedule regularly for updates. Please send us
topics and papers of your interest that could be incorporate in the
upcoming meetings. To see the papers and topics covered in the
previous year, please follow this link
for the previous acadamic year 2006/2007.
Website
last updated: September 30, 2008
by
Julia
Linke