gradient_descent_new: Gradient descent based on the out-of-sample accuracy.

In lyran92/bis557: What the Package Does (One Line, Title Case)

Description

This is a function to fit the OLS model using gradient descent that calculates the penalty based on the out-of-sample accuracy.

Usage

gradient_descent_new(
  formula,
  data_frame,
  contrasts = NULL,
  lambda = 1e-04,
  epsilon = 1e-20,
  iters = 1e+06
)

Arguments

formula	The formula of the model to be fitted.
data_frame	A data frame which contains the data for the model.
contrasts	Default is NULL. A list of contrasts for factor variables.
lambda	Default is 0.0001. The learning rate.
epsilon	Default is 1e-20. The minimum difference between the current SSR and the updated SSR.
iters	Default is 1e6. The maximum number of iterations.

Value

A list of estimated coefficients.

Examples

data(iris)
model <- gradient_descent_new(Sepal.Length ~ ., data = iris)
model$coefficients

lyran92/bis557 documentation built on Dec. 21, 2020, 10:03 p.m.