| NSF Org: |
PHY Division Of Physics |
| Recipient: |
|
| Initial Amendment Date: | June 26, 2014 |
| Latest Amendment Date: | June 26, 2014 |
| Award Number: | 1429956 |
| Award Instrument: | Standard Grant |
| Program Manager: |
Krastan Blagoev
kblagoev@nsf.gov (703)292-4666 PHY Division Of Physics MPS Directorate for Mathematical and Physical Sciences |
| Start Date: | June 15, 2014 |
| End Date: | May 31, 2015 (Estimated) |
| Total Intended Award Amount: | $16,025.00 |
| Total Awarded Amount to Date: | $16,025.00 |
| Funds Obligated to Date: |
|
| History of Investigator: |
|
| Recipient Sponsored Research Office: |
1399 HYDE PARK RD SANTA FE NM US 87501-8943 (505)946-2727 |
| Sponsor Congressional District: |
|
| Primary Place of Performance: |
NM US 87501-8943 |
| Primary Place of
Performance Congressional District: |
|
| Unique Entity Identifier (UEI): |
|
| Parent UEI: |
|
| NSF Program(s): | PHYSICS OF LIVING SYSTEMS |
| Primary Program Source: |
|
| Program Reference Code(s): |
|
| Program Element Code(s): |
|
| Award Agency Code: | 4900 |
| Fund Agency Code: | 4900 |
| Assistance Listing Number(s): | 47.049 |
ABSTRACT
During the past decade substantial advances have occurred in the area of statistical physics, both in the conceptual and analytical framework and its applications. The three-day workshop "Statistical Mechanics Foundations of Complexity - Where Do We Stand?" hosted by the Santa Fe Institute (SFI) on May 8-10, 2014 will bring together researchers from physics, computational science, biology, and network theory to discuss foundations, recent methodological developments and applications in the field of statistical physics. The aim is to bring together people with experience in a variety of novel mathematical methods with people with insights in specific complex systems. The central question of the workshop will be to critically assess the achieved theoretical progress, and to discuss the validity of the new technology in real-world applications of non-equilibrium physical, biological, etc. processes. The development of new statistical methods has an impact on problems including mathematical understanding of collective economic behavior, herding, opinion formation, financial markets, evolution, and the dynamics of innovation. The workshop is also an opportunity to encourage people outside physics to familiarize themselves with the new theoretical developments and potential applications. It will be an excellent educational opportunity for student and postdoctoral fellow participants to experience cross-disciplinary collaborations. The mixed composition of the participants will stimulate new research ideas.
PROJECT OUTCOMES REPORT
Disclaimer
This Project Outcomes Report for the General Public is displayed verbatim as submitted by the Principal Investigator (PI) for this award. Any opinions, findings, and conclusions or recommendations expressed in this Report are those of the PI and do not necessarily reflect the views of the National Science Foundation; NSF has not approved or endorsed its content.
A workshop was held at SFI May 8-10, 2014, with 30 invited participants and a number of in-house participants from the Santa Fe Institute. Over the 3 days we had 22 excellent presentations from experts in the fields of statistical mechanics, mathematics, biology, information theory, and computer science. Presentations were followed by extended discussions. The agenda of the workshop was structured in a way that pairs of speakers would either present controversial viewpoints on a given topic, or that they would mutually support each other’s viewpoints on a new development in the field. (See attached workshop agenda and participant list.)
Thematically the workshop centered around three mayor topics
- New developments in the field of generalized statistical mechanics and its use for the understanding of complex systems. This included the newly introduced class of (c,d) entropies, superstatistics, large deviation theory.
- Applications of statistical mechanics tools and information theoretical approaches to understand biological, social and ecological data. Several attempts to bridge different notions of physics and IT have been made which helped to reduce misunderstandings between the disciplinary boundaries.
- New developments in machine learning in both theory and applications. This section contained new ideas for uses of epsilon machines for new ways of the understanding the interface of information, and thermodynamics.
Fundamental questions on the applicability of generalized statistical mechanics to important model systems of complex systems. This included maybe first examples of real physical (superconducting) systems that could fully described by such a formalism.
One central set of questions discussed was which conditions have to be fulfilled to be able to talk about a true success of generalized statistical mechanics. A subset of these fundamental questions - many related to the maximum entropy principle and its generalizations - that were raised in this context, were summarized in the following way
- When does the dynamics of a system optimize entropy (or anything else)?
- Does kinetics maximize entropy even out of equilibrium?
- Does the system maximize entropy or do WE maximize it?
- Is maximum entropy the right prior for inference?
- Is maximum entropy restricted to asymptotic combinatorics, counting equally likely microstates? can we do without phase space volume and Liouville?
- How can we avoid mistakes of the past such as uncritically claiming that self-organized criticality is behind every power law?
There was broad consensus that these questions need to be answered (maybe on a system-by-system basis) before a deep understanding of the applicability and usefulness of generalized statistical mechanics can be reached.
One recurrent topic of the workshop was the question to what extent generalized statistical mechanics can be considered as "the" theory behind specific models such as unimodal maps, the Hamiltonian mean-field model, the Kuramoto model, or Pasta-Ulam like models. The general consensus seems to be that generalized statistical mechanics serves as a practical and useful effective theory for intermediate timescales, but that it is not identified as the fundamental theory behind these type of processes. Some of the presented processes are important toy cases for synchronization of agents in a system or on a network in particular.
A challenging new type of path-dependent random processes that were presented could turn out to be exact examples of generalized statistical mechanics.
One of the achievements of the workshop was that it seems that a long-standing controversy or debate could be solved: is q-statistics - and the proposal of a q-central limit theorem in particular ...
Please report errors in award information by writing to: awardsearch@nsf.gov.
