by Thalles Silva

Logistic Regression: The good parts

Everything you need to know about it.

Binary and Multiclass Logistic Regression with GD and Newton’s Method

In the last post, we tackled the problem of Machine Learning classification through the lens of dimensionality reduction. We saw how Fisher’s Linear Discriminant can project data points from higher to smaller dimensions. The projection follows two principles.

It maximizes the between-class variance.