Otto L Lecuona Sr
Local Interaction between components in complex systems
Local interactions are the engine of emergence. The behavior of the system is a function of the individual behaviorial rules of the components. Components my have different sets of rules depending on the general purpose of the system. For example, ants in a colony have different roles. Some search for food, some help build the nest, others have different functions and individual members can switch roles despending on the state of the colony or the environment. Consideration of the interconnection between components is relevent.
Components individual functions interact with other components either thru the environment or a channel for information transfer. These mechanisms are necessary for behavior modification. So I can say some characteristics of interaction are; component function, component environment interface, and information exchange.
One question comes to mind. Given a mechanism to exchange information is the channel capable of varing content? Are there differing messages in the channel? if there are then is the component able to translate that message into specific behavior?
information -> decoding->behavior determination->selection from behavioral matrix->behavior->environment/component.
Pretty straight forward process. But there are questions. Is the content of the information channel directly accessed by all components? Or is the information accepted by a particular set of components the passed forward? Is there information manipulation at each layer? Can the overall system of local information exchanged be modeled as a neural network?
I believe the model would be similiar to a neural net but not mesh connection of each set of different sets of components There could be more than one set of components behaviors. The behaviors interact locally but the different component sets are connected to different super sets. I would say agent grouping could be modeled using set theory. This is worth exploring. Common consenus is the complex systems are difficult if not impossible to model mathematically.
Leads me to think of dividing systems into deterimistic and stocastic. So we have a similiar structure as physics, classical systems and quantum. Deterministic systems could be modeled. Stochastic system are probability based. Is this a starting point for a systems theory?
To understand systems component behavior need to be understood as well their relations ships and information exchanges.