STATISTICAL-MECHANICS OF A NEUROBIOLOGICAL DYNAMIC SYSTEM - THE SPECTRUM OF LOCAL ENTROPIES (S(ALPHA)) APPLIED TO COCAINE-PERTURBED BEHAVIOR

The singularity spectrum of local entropies (S(alpha)) computed from the microcanonical dynamical partition function is used to describe the sequential dynamical structure of rat locomotor paths under the influence of cocaine. The approach used was analogous to the microcanonical ensemble of statistical mechanics and permitted the sensitive detection of "phase-transition"-like behavior. The results suggest that the behavioral sequences exhibited by the rats are determined by two processes, each with its own entropy (alpha-e, alpha-h). These processes are changed simultaneously by cocaine, suggesting that they coexist as local repellers with overlapping insets and undergo a change from second to first order phase transition.