GLMgradientMomentum: GLM Gradient descent with momentum

In damingli09/bis557: Package for homework of BIS557 Computational Statistics

Description

Fits generalized linear model by gradient descent, maximizing log-likelihood

Usage

GLMgradientMomentum(
  X,
  y,
  mu_fun,
  T_fun,
  lrate = 0.01,
  mom = 0.9,
  maxiter = 10000,
  tol = 1e-05
)

Arguments

X	the design matrix
y	the response variable
mu_fun	function from eta to the expected value
T_fun	the sufficient statistic as a function of y
lrate	the learning rate
mom	the momentum parameter
maxiter	the maximum iteration number
tol	the numerical tolerance

damingli09/bis557 documentation built on Nov. 21, 2020, 9:11 a.m.