Combining network-automata and neural networks for data analysis

Abstract

This post-doc project, to be executed into the Scientific Computing Group (SCG), aims to harness the synergy of cellular automata, machine learning, and complex networks to advance data analysis and pattern recognition techniques. The primary objective is to create novel tools for characterizing and extracting meaningful metrics from complex networks, utilizing network-automata and neural networks, to reveal latent data behavior. The project further endeavors to identify the most informative measures through pattern recognition tasks, delineating the optimal descriptors. Two key goals are the investigation and proposal of network-automata models for characterizing dynamic network evolution, and the development of techniques rooted in neural network models for learning features from complex networks and their dynamic evolution. The project's findings have vast application potential, offering substantial contributions to pattern recognition, data science, and understanding of epidemic and social processes, thereby enhancing decision-making in various domains. (AU)