Time-dependent hamiltonian functions and the representation of dynamic measurements sets

The paper describes a method for representing the dynamic clustering of time-varying data. The main goal is the extension to the continuous-time dynamic scenario of an existing method for the static case. The clustering algorithm used in the static version of the method is based on the notion of clustering function and level lines; clusters are identified as the level sets corresponding to a reference value of the clustering function. The results presented herein refer to the case in which time is introduced as an input variable of the clustering function; thus the level sets are region of the three-dimensional space and level lines become level surfaces.