Logistic-Regression

Disclosure: this post was written with the assistance of an AI system (Claude), which wrote the analysis code, ran the experiments and drafted the text. The topic, the constraints, the data choices and the final review are the author’s. Every one of 50 stratified half splits of the Wisconsin breast cancer data is linearly separable on its 30 features, so the unpenalised logistic likelihood has no maximum. scikit-learn returns coefficients of norm 2....

Disclosure: this post was written with the assistance of an AI system (Claude), which wrote the analysis code, ran the experiments and drafted the text. The topic, the constraints, the data choices and the final review are the author’s. Logistic regression has no closed form, so it is iterated, and the iteration is exactly the weighted least squares of the previous seven episodes: Newton’s method on the log-likelihood is a reweighted normal-equations solve, and it converges quadratically when there is something to converge to....