Abstract
The problem of nonparametric estimation of the intensity of a nonhomogeneous Poisson process is considered. A kernel estimator of the intensity is introduced with data driven bandwidth. The bandwidth is obtained from an L2 cross validation procedure. Results on almost sure convergence of the estimator are obtained, provided the number of observed realizations n tends to infinity. The limiting distribution of the estimator is presented for n→∞.
Similar content being viewed by others
References
Aalen, O O (1978) Nonparametric Inference for a Family of Counting Processes. Ann Statist 6: 701–726.
Boel R, Varaiya P, Wong E (1975a), Martingales on Jump processes.—I: Representation Results. SIAM Journal Control 13: 999–1021.
Boel R, Varaiya P, Wong E (1975b), Martingales on Jump Processes—II: Applications. SIAM Journal Control 13: 1022–1061.
Devroye L (1985) Nonparametric Density Estimation. The L1-view. J. Wiley & Sons New York/ Chichester/ Brisbane/ Toronto/ Singapore.
Elliot R J (1982) Stochastic Calculus and Applications, Springer-Verlag, New York/ Heidelberg/ Berlin.
Hall P, Marron S (1987) On the Amount of Noise Inherent in Bandwidth Selection for a Kernel Density Estimator. Ann Statist 15: 163–181.
Jacobsen M (1982) Statistical Analysis of Counting Processes, Springer New York/ Heidelberg/ Berlin.
Lipzer R Sh, Shiryayev A N (1986), Martingale Theory, Nauka, Moscow (in Russian).
Ramlau-Hansen H (1983) Smoothing Counting Process Intensities by Means of Kernel Functions. Ann Statist 11: 453–466.
Scott D W, Terell G R (1987), Biased and Unbiased Cross-Validation in Density Estimation. J Amer Statist Assoc 82: 1131–1146.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Schäbe, H. Nonparametric estimation of intensities of nonhomogeneous Poisson processes. Statistical Papers 34, 113–131 (1993). https://doi.org/10.1007/BF02925534
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02925534