gradient_descent: Gradient descent function for OLS regression with the penalty...

In wuyinfeistella/bis557: Linear model and the gradient descent

Description

Gradient descent is a optimization algorithm which is often used to find the minimum of a function

Usage

gradient_descent(
  formula,
  data,
  contrasts = NULL,
  alpha = 1e-04,
  tolerance = 1e-20
)

Arguments

formula	Formula of the regression model
data	The data used for gradient descent
contrasts	A list of contrasts for factor variables.
alpha	Learning rate
tolerance	The threshold of iterations

Value

The estimated coefficients for OLS regression.

Examples

data(iris)
gradient_descent(Sepal.Length ~Petal.Width, iris, alpha=0.01, contrasts = NULL, tolerance=1e-10)

wuyinfeistella/bis557 documentation built on Jan. 1, 2021, 12:52 p.m.