Kauzalitní modelování v oblasti bezpečnosti

Abstract

This Master's thesis focuses on identifying causal relationships within the field of security using the Scalable Probabilistic Approximation (SPA) method. SPA combines adaptive clustering and Bayesian modeling techniques to mathematically determine how one variable or event influences another. The goal of this thesis is to apply this method to a specific security task centered on data related to criminal activity, and to create a matrix of conditional probabilities that define causal relationships between input and output data. This approach facilitates a better understanding of risk factors and their impacts in the area.