Skip to content

Who Invented Logistic Regression?

1- Logistic Regression History

Logistic Regression was invented by Pierre François Verhulst in 1838 and he introduced it in his paper Correspondance of Mathematics and Physics which can be accessed using following link from

Verhulst was a Belgian man who was inspired and influenced by another world-changing Belgian mathematician Adolphe Quetelet.

In 1845 Verhulst went on to share an updated version of Logistic Regression he introduced in 1838 in the following paper:

  • Recherches mathématiques sur la loi d’accroissement de la population
Logistic Regression is a case of Generalized Linear Models or GLM.

2- Modern Day Applications of Logistic Regression

Logit model which is commonly used today in Logistic Regression implementations was found by McFadden in 1973 who then received a Nobel prize for this discovery in 2000. Logit model makes the assumption that error in the logistic regression is a distribution based on function of logistic density.

Logit transforms values of a line to a logistic curve by using which dependent variables (labels or target values) can only take binary values of 0 and 1.

You can find MacFadden’s original logit paper along with his other work in his personal website from Berkeley University below:

3- Summary

In this article we have covered the history of Logistic Regression, a machine learning algorithm still commonly used in 2020s, nearly 180 years after its initial discovery.