Neural Networks
Measure the prediction errors
A journey through loss functions with Botie.
Neural Networks · A journey with Botie
First visit? Learn the workshop controlsBotie's notebook
What Should Count as a Worse Answer?
Part 1 of 3
A model needs a numerical way to compare mistakes. If one prediction is slightly wrong and another is badly wrong, how much more should the second count? The answer determines which compromises the training procedure will make.