RT Journal Article
T1 Nonparametric testing of first-order structure in point processes on linear networks
A1 González Pérez, Ignacio
A1 Borrajo García, María Isabel
A1 González Manteiga, Wenceslao
AB In this paper we address a two-sample problem in the context of point processes on linear networks. The aim is to determine whether two given point patterns defined over the same linear network and under the assumption of Poissonness, share the same spatial structure. To do so, a Kolmogorov–Smirnov and a Cramér von Mises type test statistics are developed and analysed through an extensive simulation study. We have included different types of networks, balanced and unbalanced sample sizes, and homogeneous and inhomogeneous Poisson point processes. The results show a good level adjustment and high power values, the latter increasing with the sample size and the discrepancy between the two generating intensities. Finally, these methods have also been applied to the analysis of traffic accidents in Rio de Janeiro (Brazil), studying their distribution at different rush hours.
PB Springer
SN 0932-5026
YR 2025
FD 2025-01-27
LK https://hdl.handle.net/10347/39459
UL https://hdl.handle.net/10347/39459
LA eng
NO González-Pérez, I., Borrajo, M.I. & González-Manteiga, W. Nonparametric testing of first-order structure in point processes on linear networks. Stat Papers 66, 42 (2025). https://doi.org/10.1007/s00362-024-01657-8
NO María Isabel Borrajo and Wenceslao González-Manteiga acknowledge the support of from Grant PID2020-116587GB-I00 funded by MCIN/AEI/10.13039/501100011033 and the European Union. Ignacio González-Pérez thanks the Barrié Foundation, of which he is a Fellow of the 2022 call, for their support throughout his graduate studies.
DS Minerva
RD 23 abr 2026