This award supports participant travel, primarily graduate student travel, to the second AAAI Fall Symposium on Complex Adaptive Systems (CAS). The objectives for this symposium are to continue building the nascent community of complex adaptive systems researchers, and to advance understanding basic characteristics of CAS, such as resilience, robustness, and evolvability. The Symposium will address these phenomena in the context of non-trivial, real-life CAS applications, with a focus on human/social and ecological systems. Study of applications is guided by general domain-independent questions, such as:
* What are the best models for studying complex systems? * How does the structure of a complex system constrain its emergent behaviors? * What are the consequences of evolution and adaptation in complex systems? * How do we calibrate complex systems and predict their behavior?
Furthermore, communicating the properties of CAS across domains allows researchers to gain insight from fields that they normally may not encounter, to span natural, physical, social, and virtual/artificial systems.
. The main objectives for this symposium were twofold: 1) to build upon the strides made during the first AAAI Complex Adaptive Systems (CAS) Symposium, held in November of 2009, by strengthening and adding to our community of CAS researchers; and 2) to advance the next set of fundamental issues for understanding complex phenomena: namely resilience, robustness, and evolvability. Complex phenomena are characteristic of companies, societies, markets, and humans who rarely stay in a stable, predictable state for long. Yet all of these systems exhibit notable persistence of some key attributes which maintain their identities, even as their constituent parts change and adapt to new environments. What is it about these systems that define their identity? How do we characterize them, and at what scale? At what point do we distinguish between dynamic adaptability and a system-wide shift to a new phase? What are the forces that allow a system to persist, even in the face of a radically new environment? Complex Adaptive Systems have proven to be a powerful tool for exploring these and other related phenomena. We characterize CAS systems as having a significant number of self-similar agents that: Utilize one or more levels of feedback Exhibit emergent properties and self-organization Produce non-linear dynamic behavior Issues of robustness, resilience, and evolvability are critical for our understanding of complex systems, and these themes are ubiquitous across many domains. As such, our use of these concepts is a way to bring researchers together who might not otherwise find themselves in the same room. A social scientist studying population dynamics may see reflections of her work in a model of E. coli. An ecologist looking at the predator-prey relationship may find remarkably similar patterns in the immune system’s response to cancer. Or perhaps a philosopher, interested in overcoming inherent resistance to learning (change) among students, may find insight from an economist’s attempts to understand how change influences, and perhaps even destabilizes, an economic system. Encouraging the cross-pollination of ideas from one domain to another is central to our goals. As such, this year’s symposium participants – both established researchers, as well as early-career and student researchers – represent a wide variety of academic disciplines, including: Computer Science, Biology, Sociology, Physics, Political Science, Philosophy, Economics, Psychology, Network Science, Communications, Ecology, Healthcare, and Public Policy. This incredible diversity is the key to the novelty of our symposia series. This $15,000 NSF grant was used to provide funding for graduate and post-doctoral students, to partially cover the costs of attending. The symposium website can be found at the following URL: https://sites.google.com/site/complexadaptivesystems2010. Over the course of the two-and-a-half day symposium there were: 16 paper presentations, two invited speakers, an Advanced CAS modeling tutorial, a poster/demo session, and a discussion panel on Methods and Challenges for a Common Language of CAS. In addition, there were numerous opportunities for informal discussions, shared meals, and an AAAI-sponsored reception.
