Abstract
Data-driven risk networks describe many complex system dynamics arising in fields such as epidemiology and ecology. They lack explicit dynamics and have multiple sources of cost, both of which are beyond the current scope of traditional control theory. We construct the global economy risk network by combining the consensus of experts from the World Economic Forum with risk activation data to define its topology and interactions. Many of these risks, including extreme weather and drastic inflation, pose significant economic costs when active. We introduce a method for converting network interaction data into continuous dynamics to which we apply optimal control. We contribute the first method for constructing and controlling risk network dynamics based on empirically collected data. We simulate applying this method to control the spread of COVID-19 and show that the choice of risks through which the network is controlled has significant influence on both the cost of control and the total cost of keeping network stable. We additionally describe a heuristic for choosing the risks trough which the network is controlled, given a general risk network.
Highlights
Data-driven risk networks describe many complex system dynamics arising in fields such as epidemiology and ecology
We find that the seven nodes chosen incur a relatively high cost among the sample; there are many more efficient sets of control nodes one could pick. This suggests that if government intervention is enacted through these nodes, the policy decisions made in response to an infectious disease like COVID-19 would not be optimal among seven node driver sets
The dynamics resulting from these networks are difficult to analytically define for use in control theory since they must be constructed from probabilities and extensive data collection
Summary
Data-driven risk networks describe many complex system dynamics arising in fields such as epidemiology and ecology. Two problems facing the study of networked risk control are the need for explicit dynamics and optimizing over multiple cost types The former arises from the fact most real-world dynamics are informally defined by data as opposed to the formal differential equations required for control theory. These risks are active or inactive and evolve as a Markov process depending on the network’s current state. This poses a significant challenge to control theory, since applying theory from control analysis requires continuously defined dynamics for the risk variables. For a mechanical system the equations may represent the resistance of a component and the force being applied to it by other components respectively
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
