About this document

Logistic Regression Cost Function Explained by Mr. K is a document available to read on EtoBox.

The cost function for logistic regression cannot use the same function as linear regression because it would result in a wavy, non-convex function with many local optima. Instead, the logistic regression cost function is the negative log likelihood of the predicted probability matching the actual labels, calculating -log(hθ(x)) if y=1 and -log(1- hθ(x)) if y=0, where hθ(x) is the predicted probability and y is the actual label. This function results in different curve shapes when plotted based on whether th

Author

Mr. K

Language

EN