optimization: Fisher Scoring algorithm

In FREG: Functional Regression Models

Fisher Scoring algorithm

Description

Optimization algorithm for the estimation of beta regression coefficient functions and intercepts

Usage

optimization(x, y, beta, loglik, gradient, Hessian)

Arguments

x	a design matrix which is a product of inner product of basis functions and basis coefficients of functional covariate X
y	a response variable of class factor
beta	initial values for beta regression coefficients and intercepts
loglik	log-likelihood function
gradient	function for the estimation of first derivative of log-likelihood function - gradient
Hessian	function for the estimation of second derivative of log-likelihood function - Hessian

Value

beta	a vector with estimated beta regression coefficients and intercepts
ll	a value of the log-likelihood function at the estimated optimum
grd	a vector of gradient values at the estimated optimum
hessian	Hessian matrix at the estimated optimum

FREG documentation built on May 9, 2022, 5:07 p.m.