Francisco J Navarro-Meneses

Universidad Nebrija · Faculty of Social Science

After several decades of exponential growth of studies on complexity, it seems to me that society remains on the sidelines, as if all research and serious advances were considered a mere academic debate far from people's real needs. What should the complexity research community do to permeate society?

According to some authors (i.e. Stach, 2005. "Genetic learning of fuzzy cognitive maps") and my own working experience, FCMs with a large number of nodes (i.e. >15-20) and high (edge) density (80-90%), are hard to comprehend and analyze, which, besides, results in difficulties when it comes to interpreting causal- relationship simulations. What modelling technique, based on fuzzy expert knowledge, would you recommend as an alternative?

Network data analysis literature (i.e. Kolaczyk and Csardi, 2014. "Statistical Analysis of Network Data with R") suggests that comparing "real-world" network properties (degree, clustering, communities, etc.) with the same properties in random graphs (e.g. Erdos-Renyi, Watts-Strogatz) can be of particular interest given our in-depth knowledge of random graphs processes. However, it's yet hard (for me) to rise "practical" conclusions that we might infer from such comparison, other than discussing whether our "real-world" network is, or is not, different from a random graph? Any suggestion on the value-add of this analysis from a practical perspective is welcome. 

Can you suggest any idea on how to program agent-based ontologies in Matlab that can later on feed a soft computing model (i.e neural network, fuzzy logic system)? Please refer any practical example that you might know.

From a design and operational systems perspective: Is complexity always a "bad thing"? and, should simplicity be always preferred over complexity? To what extent the complexity vs simplicity debate influences systems design? Support your answer/comments with examples.

A general consensus seems to be reached on value creation involving the production and delivery of goods/services at a cost that is lower than what the consumer is willing to pay for that good/service. In accepting this approach, price does not enter the picture, since it would not affect value creation. Is this a realistic statement? If so, what is the key role of price in a value system?

As we strive to explain real-world complex systems, more parameters, variables and processes are needed in our models, thus we are less able to manage and understand our system. To overcome this "vicious circle" some authors suggest to start by defining and mapping (measuring) the complexity of the system, so as to determine a manageable degree of complexity. This involves answering the question "how much complexity is enough?". But, is this the right approach to study complexity? and if so, how may we practically accomplish these tasks?

Even though neural networks have established as an alternative to traditional statistical models, few examples can be found of real-world applications. Can you provide me with any up-to-date reference?

Following Weinberg ("An Introduction to General Systems Thinking", 1975) and Allen and Starr ("Hierarchy: Perspectives for Ecological Complexity", 1982) definition of middle-number systems as those with too many elements to be examined one-by-one and with too few elements to express them in statistical averages, do you think this could be applied to the firm?