Abstract:We investigate the nonparametric estimation problem of the density $\pi$, representing the stationary distribution of a two-dimensional system $\left(Z_t\right)_{t \in[0, T]}=\left(X_t, \lambda_t\right)_{t \in[0, T]}$. In this system, $X$ is a Hawkes-diffusion process, and $\lambda$ denotes the stochastic intensity of the Hawkes process driving the jumps of $X$. Based on the continuous observation of a path of $(X_t)$ over $[0, T]$, and initially assuming that $\lambda$ is known, we establish the convergence rate of a kernel estimator $\widehat\pi\left(x^*, y^*\right)$ of $\pi\left(x^*,y^*\right)$ as $T \rightarrow \infty$. Interestingly, this rate depends on the value of $y^*$ influenced by the baseline parameter of the Hawkes intensity process. From the rate of convergence of $\widehat\pi\left(x^*,y^*\right)$, we derive the rate of convergence for an estimator of the invariant density $\lambda$. Subsequently, we extend the study to the case where $\lambda$ is unknown, plugging an estimator of $\lambda$ in the kernel estimator and deducing new rates of convergence for the obtained estimator. The proofs establishing these convergence rates rely on probabilistic results that may hold independent interest. We introduce a Girsanov change of measure to transform the Hawkes process with intensity $\lambda$ into a Poisson process with constant intensity. To achieve this, we extend a bound for the exponential moments for the Hawkes process, originally established in the stationary case, to the non-stationary case. Lastly, we conduct a numerical study to illustrate the obtained rates of convergence of our estimators.

Note on the approximation of the conditional intensity of non-stationary cluster point processes

Approximation intensity for pairwise interaction Gibbs point processes using determinantal point processes

Nonparametric Estimation of Quantiles for A Class of Stationary Processes

Strong Approximation for A Class of Stationary Processes

Poisson intensity parameter estimation for stationary Gibbs point processes of finite interaction range

Predicting the intensity of partially observed data from a revisited kriging for point processes

Nonparametric estimation of the stationary density for Hawkes-diffusion systems with known and unknown intensity

Moment formulae for general point processes

New density/likelihood representations for Gibbs models based on generating functionals of point processes

Nonparametric intensity estimation from noisy observations of a Poisson process under unknown error distribution

Conditional Poisson process approximation

Likelihood inference of the non-stationary Hawkes process with non-exponential kernel

Stein estimation of the intensity of a spatial homogeneous Poisson point process

Adaptive estimating function inference for non-stationary determinantal point processes

Erlang mixture modeling for Poisson process intensities

Approximation with ergodic processes and testability

Lyapunov criteria for uniform convergence of conditional distributions of absorbed Markov processes

Approximation Rates of Nearest Neighbor-Kernel Estimator for the Conditional Density

Weak approximation rates for integral functionals of Markov processes

Minimum contrast for the first-order intensity estimation of spatial and spatio-temporal point processes

Poisson limit theorem for the number of excursions above high and medium levels by Gaussian stationary sequences