Actions

CSSS 2008 Santa Fe-Tutorials: Difference between revisions

From Santa Fe Institute Events Wiki

Line 33: Line 33:
== Introduction to the Design and Analysis of Computer Experiments ==
== Introduction to the Design and Analysis of Computer Experiments ==
How do you find "interesting" behavior when your simulator is too slow or the inputs are too many to try every possible combination? Using an Arctic sea ice simulator, I will show you how modern statistical methods can help you explore your virtual world more efficiently. Click  
How do you find "interesting" behavior when your simulator is too slow or the inputs are too many to try every possible combination? Using an Arctic sea ice simulator, I will show you how modern statistical methods can help you explore your virtual world more efficiently. Click  
http://www.stat.sfu.ca/~dbingham/NICDS_CompExpt/research.html here for more details] or  
[http://www.stat.sfu.ca/~dbingham/NICDS_CompExpt/research.html here for more details] or  
[http://www.schonlau.net/publication/jogo98.pdf here for a paper about application to global optimization].
[http://www.schonlau.net/publication/jogo98.pdf here for a paper about application to global optimization].
Time permitting we could also touch on some of the statistical concepts involved, e.g. cross-validation, maximum likelihood estimation, or Bayesian statistics.
Time permitting we could also touch on some of the statistical concepts involved, e.g. cross-validation, maximum likelihood estimation, or Bayesian statistics.
[[B%C3%A9la_Nagy|Béla]]
[[B%C3%A9la_Nagy|Béla]]

Revision as of 05:08, 3 June 2008

Template:CSSS 2008 Santa Fe


Open Source research software

Open (no pun intended...) your eyes to the wonderful world of FOSS -- Free and Open Source Software. While the distinction between Free and Open Source is a very interesting one (and highly contentious in the right crowds), for research purposes, we want to use the best tools for the job but some of us suffer from limited income, so I would like to talk about both. At the same time, FOSS is great to use for a non-economic reason: if you find bugs, or design extensions, you can fix them yourself (in OSS, at least) or at least report the problem back to a typically active community. Some potential tools to discuss/explore: R, Octave, Scilab, Gnuplot, perhaps some of the more useful languages in the field like Python and perhaps others I don't know as much about (a quick `apt-cache` on my Ubuntu Hardy install shows RasMol, ClustalW, SeaView, Achilles, complearn, EMBOSS, GENESIS, etc...)

R tutorial

I (Ruben) know a little bit about R (basic stuff such as common plots and regression analysis) but would like to enhance my R skills. Does anybody have an interest in a R tutorial, too? Please edit this if there is more interest.

Python tutorial

I've (also Ruben) interest in a Python tutorial. Please edit this if there is more interest.

I can give a tutorial on python and on scipy/numpy, its great scientific library Giovanni

Neutral models in biology

There is an interesting paper by Cosma Shalizi of SFI about methodological problems in social sciences research in which he talks about the concept of neutral models in evolution models. I was wondering if any of the bio-people can give a tutorial on this topic as I am pretty interested in understanding the concept. Giovanni

GIS

I (Flavia) could offer an introduction to GIS/Spatial Analysis. Please edit here if you are interested.

Statistical Physics for Non-Physicists

Problem: Textbooks about this are writen for physicists. Solution: A Physicist (or mathematician) that would be so kind and spend few minutes (or maybe hours) to explain all that stuff to people like me(Petr):-)

Ruben: Do you seek for a general introduction or something specific?

Modern Logic and Reasoning

Like I mentioned in the 'ice-breaking', I could tell something about application of modern logic into human reasoning. It's a very board topic, and very new. Criticisms are welcome and needed. I would give some basic examples. On top of that, I would also say some development of logic, and how I found it useful in research, which might seem un-related to logic, esp. in social science. I am planning to give a 15 to 20 minutes presentation, UNLESS people want to hear more, in that case, please let me know. QiQi

Introduction to the Design and Analysis of Computer Experiments

How do you find "interesting" behavior when your simulator is too slow or the inputs are too many to try every possible combination? Using an Arctic sea ice simulator, I will show you how modern statistical methods can help you explore your virtual world more efficiently. Click here for more details or here for a paper about application to global optimization. Time permitting we could also touch on some of the statistical concepts involved, e.g. cross-validation, maximum likelihood estimation, or Bayesian statistics. Béla