CSSS 2008 Santa Fe-Tutorials
From Santa Fe Institute Events Wiki
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...)
I would be happy to contribute a little bit about freely available simulation environments like ns-2 (computer networks) and omnet++ (a generic DES) -- Laura
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.
Would be interested in learning about this as well. -Devin
I would be interested as well. Mark
I am fairly familiar with R and could probably run a tutorial... what are you all interested in learn? - Skyler
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. I can also talk about coding in general, as python is both a languange which is object oriented, imperative and functional (somehow). We can use the python tutorial itself as a reference for the part about the language, and then move to the basic concepts of the duo numpy / scipy, which form a powerful tool to manipulate n-dimensional arrays of numbers and also talk about ipython (the enhanced interactive shell) and the pylab interface, which gives a very nice environment for interactive programming and data analysis. Since pylab has been designed to mimic MATLAB's interface (the major plotting/statistical functions work as expected in both enviroments, which saves you a lot of time if you're used to MATLAB), I can also talk a bit about MATLAB, but being not a big fan of it, it would be better if somebody else stepped in to another tutorial on that.
Leave a mark if interested! 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
I would be very interested in learning about neutral networks too! - Skyler
GIS / Spatial Analysis
Space Matters!! Geographical information system (GIS) is a computational system (hardware + database engine) that is designed to assemble, store, update, analyze, manipulate, and display geographically referenced information (data identified by their locations).
I'm thinking about introducing some basic GIS concepts and a free GIS software known as Terraview. We could also explore some spatial analysis techniques (this is the best part!) using Terraview and GeoDa (also free!).
Please edit here if you are interested or send me an email. Flávia
Statistical Physics for Non-Physicists
Problem: Textbooks about this are written 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?
A crash course in Statistical Physics would be awesome. Let us know. — Carlos
I am interested too. (Soumya)
Me too. (Jean)
Ditto - Skyler
I'd be willing to run such a tutorial. However, I would have to consult with some/all of the interested parties to find out what kind of statphys you want to learn about. There are a huge number of possible topics, one could start with basics like ensembles, or perhaps people are interested in master equation and other non-equilibrium techniques, or maybe critical phenomena is what people are interested in. I really do not know. (Orion)
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
Update! Time at June 10th, 03.15 p.m. - 03.35 p.m. CSSS 2008 Santa Fe-Schedule
Introduction to the Design and Analysis of Computer Experiments
How do you find "interesting" behavior when your computer model is too slow or the inputs are too many to try every possible combination? Using an Arctic sea ice simulator example, I will show you how modern statistical methods can help you explore your virtual world more efficiently. Check out this brief overview or a more technical paper about 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
Genomics / Central dogma overview
It seems like some of you might be interested in an overview of the central dogma of molecular biology to non-biologists. This could be an 1h tutorial on the major actors of gene expression: nucleus, chromosomes, chromatine, DNA, RNA (tRNA, mRNA), proteins, polymerases, ribosomes, transcription factors, and eventually a quick intro to small, non-coding RNAs as a bonus. Although being a bioinformatician by training, I'm happy to leave the way if a "hard core" biologist wants to do this tutorial (Molly ?). Edit if interested! Jean
I'd be happy to attend a tutorial in 'genomics for idiots' -- Laura
Me too. I am also interested in metagenomics if this is not too much of course. -- Francois
Artificial Intelligence/Machine Learning
I (Nish) could introduce some of the basic methods in AI/ML. If there is significant interest in the two fields separately, I could do two tutorials. Would probably focus on the higher level, rather than the nitty-gritty details, as well as applications of the methods to real problems. I'm not necessarily an expert, although have a fair amount of experience in the area, so I would prefer a more interactive session, where questions can be answered by everyone.
Sign me up. — Carlos
A Crash Course to Classical and Evolutionary Game Theory
Game theory is the study of interactive decision making. Classical game theory aims to develop a general theory to describe how rational agents interact strategically. In many cases humans lack the kind of infinite computational power and time assumed by classical game theory. In the early 1970s the biologist John Maynard Smith introduced evolutionary methods to the field, dispensing with the assumption of hyper-rationality while changing many of the concepts central to the field along the way. The result was evolutionary game theory. This new framework has been used to model the behavior of fundamentally non-rational players (such as viruses) as well as humans.
In this tutorial, I'd try to introduce the basic concepts in both of these fields, namely, the definition of a game, payoffs, the Nash equilibirum and evolutionarily stable strategies, the replicator dynamics. I'll briefly mention the three basic classes of two-strategy games represented by the Prisoner's Dilemma, the Snowdrift Game (sometimes called the Hawk-Dove game or Chicken), and the Stag Hunt Game. Depending on particular interests of the group, we could prove the Bishop-Cannings theorem and give a classification of all symmetric two-strategy games; or look at updating methods and spatial chaos; reputation and image scoring; rock-paper-scissors in biological systems; or evolutionary branching and specialization.
If there's something else you'd like to know about EGT, shoot me (Josh) an email, and I'll see if I can dig up something I know on your topic. I'm not going to require any fancy mathematical background. If you've seen a 2×2 matrix before, great. Otherwise, it's not a big deal. We won't multiply them or calculate their eigenvalues. They'll just serve as a means for bookkeeping.
Update! Some of us are also thinking about setting up a working group as well.
Update 2! This has been scheduled on Friday from 3 - 5, location TBD.
- I'll sign up for this. Kolbjørn
- I'm interested too! (Flavia)
- I'll be there too. Kathleen
- I'm in. Jean
- I'm interested as well. Steve
- will be there at 3 Walt
- I'll be there, Petr
The resilience perspective is increasingly used as an approach for understanding the dynamics of social–ecological systems. Essential for the resilience perspective is the recognition that living systems are not in equilibrium but rather in a domain of attraction. Many dynamic systems, however, have multiple domains of attraction. Moreover, self-organizing processes can create or change the shape and depth of this domain of attraction. Within the resilience perspective, new pathways of sustainable development can be represented by crossing a threshold from a domain of attraction and/or by creating new domains. Resilience is a measure of how much change or disruption is required to transform a system from being maintained by one set of mutually reinforcing processes and structures to a different set of processes and structures. If you are interested we (Mike and Dirk) can introduce you to some of the insights developed by the resiliance alliance and the challenges we face in understanding these kind of systems.
I am interested in this too. Richard
Very interested, any idea of when you will do it? Walt
I'm interested as well. Steve
I'm interested too. - Skyler
Introduction to classical control theory
I (Srideep) can offer a 'quick' tutorial on control theory/control systems. This is will be a simple introduction to the motivation, basic ideas, issues and jargon in the field. If you are interested, please let me know about your background in linear algebra, complex analysis and calculus. Depending on the background, I might spend more or less time introducing the field.
Ideally, if you know what eigenvalues and eigenvectors of a matrix are, what a pole of a complex function is and how the solution of a linear differential equation looks like, you are ready to jump right into controls. If the words above don't mean much at all, then we can run a quick 'review' of what they mean intuitively. you can sign up here or send me an email srideep
I'd be very interested in this tutorial. I think I'm basically OK on the prerequisites, but I wouldn't be annoyed by a review. Perhaps Monday? -- Laura
I'm in too. I guess I should be ok on linear algebra, calculus and linear ODEs, but I don't know what the pole of a complex function is. Jean
I'm in. If we can start with 'pole' thing, that would be wonderful. - Masayoshi
Hi Srideep, please put me in this group. About my background on the subjects you asked; zero!!! Sorry. Rio
I (Srideep) will also be happy to talk about topology, introducing the concepts of point-set topology. The language of modern mathematics is enshrined in the concepts of point-set topology. I can also talk about group theory and introduce abstract algebra to those interested. In my opinion, it is the most powerful gateway into abstract thinking. sign here or email me srideep
I'd be very interested in that. Jean
I'd be interested to see what you cover in the topology section. Algebra, however, is for the birds :) Paul
Eigenvalues - what are they and how to find them?
I (Kolbjørn) can put together a brief and elementary introduction to eigenvalues and eigenvectors if anyone have an urge for this. Sign up or e-mail and we'll schedule something. Kolbjørn
Yes please. Kathleen
I am also very interested Walt
I'm in as well Mark
Please - have always been kind of confusing to me. Jon
How your computer works
Nish and Laura can give a joint tutorial on 'how your computer works'. What happens when I type 'www.santafe.edu' in my browser? How does a web server at santafe.edu handle all those incoming requests? What happens when I use a WiFi access point? Basically, we'd be happy to take your questions about how your computer works and do our best to answer them - we're also happy to have other co-tutors.
Let us know if there's interest Laura, we'd probably schedule later next week, to not conflict with tutorials that focus on maths and other project prerequisites.
How your hardware works
Along the same lines as the computer tutorial, I've found myself discussing hardware with a number of folks. And why hardware matters from a massive parallelism perspective (which is quite common in the complex research areas I've encountered). If folks are interested, I can give a rough overview of the way hardware works in different types of computers and supercomputers (as much as I understand of it) as well as how to best leverage that knowledge.
Computational Physics for Non-physicists or A small introduction into Applied Physics
I've seen that many people are interested in physics. I could give an introduction to "computational" physics - this means physics with a PC. Actually, it is very broad and gives some basics for simulations (interesting for all simulation-folks):
- What is a 'random number generator' and why should I know something about it?
- What are Master-equations?
- The Ising-model / Voter-model
- The Central Limit Theorem or why does it make sense to average over multiple runs of a simulation?
I'm very interested Nish
An open discussion of Shannon information theory (would like some help in presenting this part clearly) and then some newer results from its application to cellular automata (and potentially other complex systems).
CAs (particularly ECAs) are a very interested model of computation. How do 8 rules (ECA 110, e.g.) emulate a Turing Machine? Why is that interesting? What can we learn about what defines computation given CAs? Maybe we can also discuss some simple computational (Turing) theory.