logistic_regression: Logistic Regression

In AlbertoAlmuinha/LogisticRegression: Logistic Regression Implementation for Binary Classification Problems

Description

Logistic Regression for Binary Classification Problems.

Usage

logistic_regression(train, test, target, lr = 0.001, max_iter = 1e+05,
  probs = F, threshold = NULL, regu_factor = NULL,
  regu_method = "ridge", epsilon = 1e-05, ...)

Arguments

train	Matrix with the predictor variables
test	Matrix with the target variable
target	Numeric value of the target variable
lr	Learning Rate for Gradient Descent Algorithm
max_iter	Number of iterations for Gradient Descent Algorithm
probs	If the result must be given in probabilities or not.
threshold	Threshold for gradient descent algorithm
regu_factor	Regularization Factor for Ridge Regularization Regression
regu_method	Regularization Method. At this moment, only "ridge" available
epsilon	Numeric value
...	Other possible arguments

Value

Returns a list with the predicted data, confussion matrix, confussion matrix plot and optimal theta values

Author(s)

Alberto Almuiña

AlbertoAlmuinha/LogisticRegression documentation built on Jan. 5, 2020, 9:18 a.m.