John Harte - Disturbance and Recovery Dynamics in Complex Systems
27:08 "Different mechanisms for different disturbances. So we have to include explicit mechanisms if we are to understand disturbed ecosystems." This is exactly what Jaynes did in his Where do we stand on Maximum Entropy? essay, isn't it? He gave the example of Wolf's dice data (pages 48-58), which was some data from actual throws of a real physical die which had a slight deviation from a uniform distribution. He then described a physical model of the important parameters and from an hypothesis as to how the die had been made he proposed some parameters, variations of which would have produced some changes in the distribution. Applying the MAXENT hypothesis to the model with the unknown perturbations he was able to show that the model produced a best fit.
But in the case of a recovering ecosystem one presumably has a higher-order problem because the dynamics of the ecosystem would change as the populations changed.
John Harte is in his late eighties and he still has a job!
Subscribe to Sants Fe Institute.
I wonder if anyone has argued that the success of Maximum Entropy methods in modelling complex systems is evidence of adaptation and learning in those systems? See James Corbett on Cybernetics and in particular this short remark of Stafford Beer:
Subscribe to Josh Moore.
Comments
Post a Comment