Module pointprocesses::estimators::nadarayawatson

Implement the Nadaraya-Watson non-parametric regression estimator. Useful for estimating the intensity of a non-homogeneous Poisson process.

Structs

NadWatEstimator	Nadaraya-Watson nonparametric estimator for functions using a weighted kernel average. The predictor at a point $x_0$ is given by: $$ \hat y_0 = \frac{\sum_{i=1}^p K_h(x_i, x_0) y_i} {\sum_{i=1}^p K_h(x_i, x_0)} $$
SmoothingKernelIntensity	Estimate the intensity function of an event sequence under a variable Poisson model using a kernel smoother (see A kernel method for smoothing point process data by P. Diggle). The regressor is given by $$ \hat\lambda(t) = e_h(t)^{-1} \sum_i K_h(t - t_i) $$ where $e_h(t) = \int_0^T K_h(t - u)\, du$ is an edge-correction term.

Type Definitions

UniformKernelIntensity

Intensity kernel estimator using a uniform kernel.