Friday, January 29, 2010

Slime Mold and rail links in Japan

Maybe a bit off topic but recently in the NY Times there was an interesting article that highlighted researchers in Japan have used slime mold (single-celled amoeboid organism) to grow networks which show high correspondence to the rail network in the area. around Tokyo. They did this by placing 36 bits of food in a pattern corresponding to cities in the Tokyo area and letting the slime mold grow from a spot corresponding to Tokyo.

The abstract for those interested in Science is:

"Transport networks are ubiquitous in both social and biological systems. Robust network performance involves a complex trade-off involving cost, transport efficiency, and fault tolerance. Biological networks have been honed by many cycles of evolutionary selection pressure and are likely to yield reasonable solutions to such combinatorial optimization problems. Furthermore, they develop without centralized control and may represent a readily scalable solution for growing networks in general. We show that the slime mold Physarum polycephalum forms networks with comparable efficiency, fault tolerance, and cost to those of real-world infrastructure networks—in this case, the Tokyo rail system. The core mechanisms needed for adaptive network formation can be captured in a biologically inspired mathematical model that may be useful to guide network construction in other domains."

Science 22 January 2010: Vol. 327. no. 5964, pp. 439 - 442

Read the full report in Science or in the NY Times

Thursday, January 14, 2010

Ph.D Program in Computational Social Sciences

This might be of interest to some readers of the blog. Its a Ph.D program offered by the Department of Computational Social Science at George Mason University.

George Mason University - Ph.D Program in Computational Social Sciences

The core objective of the Ph.D. program in Computational Social Science is to train graduate students to be professional computational social scientists in academia, government or business. Our program offers students a unique and innovative interdisciplinary academic environment for systematically exploring, discovering, and developing their skills to successfully follow careers in one of the areas of computational social science.

Examples of areas of concentrations and potential specializations include but are not limited to the following:
  • Agent-based computational economics: trade, finance, decision-making under risk
  • Computational political economy: voting, institutions, norms, inequality
  • Computational linguistics: generative grammars, parsing, classifiers, inference
  • Social network analysis: connectivity, structure, evolution of the WWW, cyberwarfare
  • Computational anthropology: emergence of hierarchy, settlement patterns
  • Computational political science: systems of government, conflict and war, cooperation
  • Computational sociology: segregation, collective action, leadership, trust
  • Complexity theory: power laws, potential theory, criticality, bifurcation
  • Computational methodology: multi-agent systems, evolutionary computation, UML, GIS, visualization, sonification, computational epistemology

Admission Requirements and Procedures

The application deadline is February 1st of each year for students seeking financial aid, or April 1 for all other students.

Applicants should have as background a bachelor's degree in either one of the social sciences, in computer science, engineering, or in a relevant discipline, as well as undergraduate courses in these and related areas. Bachelor's degrees in the physical or biological sciences are also eligible, but applicant may be advised to take additional courses in social science or computer science as prerequisites to admission. Minimal requirements also include one undergraduate course in calculus and knowledge of a computer programming language preferably object-based. While in the program students are expected to develop significant expertise in the utilization of computational social science resources such as agent-based simulations or other computational tools. The program maintains a simulation environment, the Multi-Agent Simulator of Neighborhoods and Networks (MASON), in collaboration with the Evolutionary Computation Laboratory (EC Lab) of the Department of Computer Science. Mathematics training beyond basic calculus is not required, but may be useful in some areas of specialization.

For more information, please see http://www.css.gmu.edu/

Online applications can be submitted to http://admissions.gmu.edu/grad/.

Wednesday, January 06, 2010

Dynamic Modeling in a GIS Environment Seminars Online

Just a quick post to highlight that the presentations from the "Dynamic Modeling in a GIS Environment" Virtual Seminars from the Global GIS Academy are now availble online. For those interested in GIS and various aspects of modelling these seminars give a glimpse into some current research.

Also if you like to look at online seminars the 2008 Autumn seminar series on Neogeography is also available.

Monday, January 04, 2010

modelling4All update

I been following the progress of modelling4All for a while now and thought it worth highlighting again (click here to see an earlier post), especially after a post on the modelling4All blog noting some improvements.

The aim of modelling4All is to reduce the difficulty and effort needed to make agent-based models therefore enabling non-programmers to collaboratively build and analyze computer models. modelling4All allows one to create agent-based models over the web, models are constructed by composing and parametizing model fragments (bits of code) as shown in the movie below.



The model fragment library can be searched and model fragments can be added and combined to a model to create an agent-based model in a very short period of time by just clicking on the fragment (such as a behavior) which is required for the task at hand. Click here to see such a model, in this case its an epidemic model where people go to work, school, and home. A neat feature is also the ability to export the models as NetLogo models.