Sporala red del conocimiento
EN
Universidad Nacional de Colombia

Medellín - Ciencias - Doctorado en Ciencias - Estadística · 2025

Methods for features classification in point processes on linear networks

In this work, we propose two methods for the analysis of point processes on linear networks with different approaches. We extend methods developed in previous studies to this more complex geometric context, where the classical properties of a point process change and data visualization are not intuitive. The first method is for detecting clusters of points within clustered point patterns in linear networks, focusing on the classification of point processes. Our approach leverages the distribution of the K-th nearest neighbour volumes in linear networks. Our methodology is well-suited for analysing point patterns comprising two overlapping Poisson processes occurring on the same linear network. The second method consider the problem of testing the hypothesis of complete spatial randomness in homogeneous point processes on linear networks. We propose a statistical test based on the fractal dimension, calculated through the Box-counting method. As a result, the method is suitable for testing if a point pattern on linear network is completely random (uniform Poisson) and also to discriminate between clustered or inhibitory behaviour of the point pattern. We present simulations and examples to illustrate these methods. (Tomado de la fuente)

Texto completo 81 páginas con texto

Leer la tesis completa Ficha en el repositorio

Tesauro de su biblioteca