14 February 2017

A.I. detecting early-warning signals …in the complexity of Digital Society and Economy

The Digital Society and Economy are literally becoming complex systems. All sub-systems, at the different levels, are hyper-connected with non linear relationships. Sudden regime/phase transition can occur radically changing the scenarios
Can we early detect tipping points of such sudden regime/phase transition ?

Predicting such tipping points before they are reached is quite difficult, but it might have a huge impact in several fields, from medicine to business, from biology to meteorology to social networking, from management to business and to cyber-security.

There are some nice research papers providing some guidelines. This is one of them “Early-warning signals for critical transitions”. In this case, the paper is suggesting the analysis of generic early-warning signals indicating, for a wide class of complex systems, the approaching of a critical threshold, where small forces can cause major changes in the state. Examples of such transitions might include the collapse of over-harvested ecosystems, climatic changes, or stocks markets dynamics.

For example one symptom is the critical slowing down: when the system approaches a critical transition, it becomes increasingly slow in recovering from small perturbations (which is translated mathematically into an increase in the autocorrelation and variance of the fluctuations). Another signal that can be seen in the vicinity of a catastrophic transition point is flickering. Stochastically, the system moves back and forth between the basins of attraction of two alternative attractors (bistable region). Spatial patterns is a third example: an ecosystem may show a predictable sequence of self-organized spatial patterns as it approach a critical transition (e.g. a semi-arid vegetation to increasing dryness of the climate).

Another recommended reading is this one:


Big Data analysis by A.I. systems could make a breakthrough in this promising area of research and innovation, also on the path towards a sustainable 5G.
The potential gains of investing in these studies are formidable.