Application and development of advanced mathematical tools for population and time series analysis in pulsar astrophysics

Abstract

In this thesis, we introduce novel methods for analyzing pulsar populations using a variety of mathematical techniques. These tools-particularly graph theory-have been thoroughly validated in advanced mathematics, enabling us to overcome some of the constraints (even dimensional) inherent in conventional visualization approaches. This exploration benefits from dimensionality reduction techniques, which not only lessen computational demands but also highlight potential for describing physical characteristics. The resulting structures encode information about pulsar similarities that extend beyond standard spin parameters, revealing relationships that are not readily apparent in traditional diagrams. With a physically motivated topological perspective, we leverage the strengths of these methods and present results that span from prospective source classification and the emergence of new classes to catalog comparison, among other applications. This new approach enables fresh interpretations of longstanding problems, laying a new foundation for visualizing the pulsar population and categorizing sources. Building on this, we identify several sources as likely members of specific binary subclasses and investigate the potential transitional nature of others. Furthermore, we extend the use of graph theory to the boundary of machine learning, demonstrating its capability for binary separation in an unsupervised context. Finally, we introduce and apply an innovative, flexible time-series alignment technique to the field of gamma-ray astrophysics. The method identifies notable similarities among the light curves of gamma-ray pulsars. The results presented here are promising, offering a refreshing direction for the field and new pathways for rigorous mathematical analysis, ultimately providing meaningful alternatives to traditional approaches in high-energy astrophysics.