Abstract
This paper presents some innovative techniques for the analysis and prediction of diffusion phenomena: The Target Diffusion Model (TDM) and the Twisting Theory (TWT).TDM is an algorithm able to reconstruct the global causation process of diffusion in space and time with a bottom-up data driven approach, while TWT is an algorithm based on twisting theory and able to trace trajectories of any unmapped entity belonging to the same diffusion space as a consequence of changes in the original fixed entities. We test these techniques on both an illustrative problem and a real-world case, specifically, the 2001 Dengue fever epidemic outbreak in Brazil. This approach obtains interesting performance enhancements in both analytical results and predictive capabilities. These results appear to be especially useful in applications for which an inherent structural grammar of complexity can be found. We also carry out analyses by means of static analytical tools such as the Topological Weighted Centroid (TWC) and the Harmonic Center and we find that, overall, the whole suite of techniques is able to provide a fine-grained characterization of the main features of the diffusion process. We consider our approach to be a promising step for extracting relationships, prediction, and the understanding of the dynamics of spatiotemporal phenomena. This development is a meta-disciplinary perspective to the study of diffusion phenomena.
Published Version
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