fit: Fit Function

In FD155/ProjetR: Logistic Regression

Description

Implementation of binary logistic regression with stochastic gradient descent for algorithm

Usage

fit(
  formula,
  data,
  mode,
  batch_size,
  ncores = 0,
  coef,
  max_iter = 100,
  learningrate = 0.1,
  tol = 1e-04
)

Arguments

formula	Target variable ~ explanatory variables
data	dataframe required to perform the gradient descent
mode	Gradient descent mode (Online, Mini-Batch, Batch)
batch_size	size of the cutting to be carried out for the "mini-batch" mode
ncores	Number of cores required for parallelization
coef	Fixed number for creating the base vector
max_iter	Max number of iterations not to be exceeded
learningrate	Learning rate which determines the step of the descent
tol	Tolerance threshold for which convergence is said to be "finished"

FD155/ProjetR documentation built on Dec. 17, 2021, 7:31 p.m.