English

Adaptive wavelet estimation of a compound Poisson process

Statistics Theory 2012-03-15 v1 Statistics Theory

Abstract

We study the nonparametric estimation of the jump density of a compound Poisson process from the discrete observation of one trajectory over [0,T][0,T]. We consider the microscopic regime when the sampling rate Δ=ΔT0\Delta=\Delta_T\rightarrow0 as TT\rightarrow\infty. We propose an adaptive wavelet threshold density estimator and study its performance for the LpL_p loss, p1p\geq 1, over Besov spaces. The main novelty is that we achieve minimax rates of convergence for sampling rates ΔT\Delta_T that vanish with TT at arbitrary polynomial rates. More precicely, our estimator attains minimax rates of convergence provided there exists a constant K1K\geq 1 such that the sampling rate ΔT\Delta_T satisfies TΔT2K+21.T\Delta_T^{2K+2}\leq 1. If this condition cannot be satisfied we still provide an upper bound for our estimator. The estimating procedure is based on the inversion of the compounding operator in the same spirit as Buchmann and Gr\"ubel (2003).

Keywords

Cite

@article{arxiv.1203.3135,
  title  = {Adaptive wavelet estimation of a compound Poisson process},
  author = {Céline Duval},
  journal= {arXiv preprint arXiv:1203.3135},
  year   = {2012}
}
R2 v1 2026-06-21T20:34:01.314Z