10-03

Cross-entropy loss and gradient descent

Today we talked about how the weights and bias in a logistic regression model are actually learned. We talked about the gradient descent algorithm, and the cross-entropy loss function, which quantifies error. Please read sections 5.4 through 5.6 of Chapter 5 in the Jurafsky and Martin textbook. Specifically, pay attention to pages 11-17.

Link to Chapter 5 here.

Link to the full book here.