Logistic regression

Logistic regression estimates the probability of a categorical outcome by applying a logistic function to a linear combination of predictor variables. Its coefficients describe how predictors change the outcome’s log-odds.

Logistic function: A sigmoid function that maps real-valued inputs to values between zero and one. It transforms the model’s linear score into an estimated probability.
Categorical variable: A variable whose values represent distinct categories rather than measured quantities. Logistic regression models categorical outcomes, often binary ones.
Multinomial logistic regression: A logistic regression model for outcomes with more than two unordered categories. It extends probability modeling beyond binary outcomes.
Linear regression: A regression method that models an outcome as a linear combination of predictors. Unlike logistic regression, it can predict values outside the probability range.
Logit: The logarithm of a probability divided by its complement. The model treats outcome log-odds as a linear function of predictors.
Binary classification: A prediction task that assigns cases to one of two classes. Binary logistic regression is the standard form for two-outcome prediction.
Ordinal logistic regression: A regression model for ordered categorical outcomes that estimates cumulative category probabilities. It uses outcome ordering when categories have a meaningful rank.
Probit model: A binary-outcome regression model that applies the standard normal cumulative distribution function to a linear predictor. It replaces the logistic link with a normal-distribution link.
Maximum likelihood estimation: A method for estimating parameters by maximizing the probability of observed data under a model. It commonly fits logistic regression coefficients to observed outcomes.
Probability: A numerical measure from zero to one representing the likelihood of an event. The model’s central prediction is an outcome probability.

Linked from 13 pages