parse_kernel_variable

(vector of characters) Names of variables in data to
create the kernel matrix from. Must be a single term that is either of the form
\(x\) (a single linear term) or \(k(x1, x2, \dots)\) (a kernel term that may
contain multiple variables).

kern_var_name

(function) A kernel function. Will be overwritten to linear
kernel if the variable doesn't contain 'k()'.

kern_func

(data.frame, n*d) A data.frame, list or environment (or object
coercible by as.data.frame to a data.frame), containing the variables in
formula. Neither a matrix nor an array will be accepted.

data

(data.frame, n_new*d) New data for computing predictions.

data_new

Create kernel matrix for each variable in the formula.

internal

Implementation of Cross-Validated Kernel Ensemble (CVEK),
a flexible modeling framework for robust nonlinear regression and
hypothesis testing based on ensemble learning with kernel-ridge estimators
(Jeremiah et al. (2017) <arXiv:1710.01406> and
Wenying et al. (2018) <arXiv:1811.11025>). It allows user to conduct
nonlinear regression with minimal assumption on the function form by
aggregating nonlinear models generated from a diverse collection of kernel
families. It also provides utilities to test for the estimated nonlinear
effect under this ensemble estimator, using either the asymptotic or
the bootstrap version of a generalized score test.

parse_kernel_variable: Create Kernel Matrix

Description

Usage

Arguments

Value