3 ms·
Learn about information theory. It is better to calculate the cross-entropy error or KL-divergence between the data and the model.
by jostmey 10y ago
Learn about information theory. It is better to calculate the cross-entropy error or KL-divergence between the data and the model.
- chestervonwinch 10y agoDo you have any links where KL-divergence is used in the case of real-valued response variables? Why is it better?