Question
What are the best research groups investigating complex systems and complex networks?
Answer
Besides Santa Fe, there are excellent complexity programs at U. Michigan and Cornell, and something called the New England Complexity Research Institute.
http://necsi.edu
But really, it is less of a discrete community today than a set of techniques and research approaches that have spread all over STEM academia. You can learn and do "complexity flavored" work just about anywhere today. The field was a temporary fork that has now re-merged into the mainstream. The descendant influence is most evident today in fields like machine learning, mathematical sociology, social data analysis, systems biology, distributed AI, distributed control, UAV research, driverless cars, formation flight, behavioral finance etc.
The field itself is really a historical field now. I'd say it peaked as an independent field with the complex networks work on 1997-2003 (Barabasi, Watts etc.) and has produced no comparable body of big ideas since.
I'd say the decline was mainly driven by availability of huge data sets. It became more interesting to look at the data and model what it actually revealed than explore theories based on the characteristic "simple models can yield complex behavior" aesthetic of classic complexity research.
In other words, our attention shifted from "toy" complexity in simulations to the real thing.
http://necsi.edu
But really, it is less of a discrete community today than a set of techniques and research approaches that have spread all over STEM academia. You can learn and do "complexity flavored" work just about anywhere today. The field was a temporary fork that has now re-merged into the mainstream. The descendant influence is most evident today in fields like machine learning, mathematical sociology, social data analysis, systems biology, distributed AI, distributed control, UAV research, driverless cars, formation flight, behavioral finance etc.
The field itself is really a historical field now. I'd say it peaked as an independent field with the complex networks work on 1997-2003 (Barabasi, Watts etc.) and has produced no comparable body of big ideas since.
I'd say the decline was mainly driven by availability of huge data sets. It became more interesting to look at the data and model what it actually revealed than explore theories based on the characteristic "simple models can yield complex behavior" aesthetic of classic complexity research.
In other words, our attention shifted from "toy" complexity in simulations to the real thing.