Loss Functions

# Loss Functions: Measuring The Error Of Predictions Imagine training a puppy. You tell it to sit, and sometimes it does, sometimes it rolls over, and sometimes it just stares blankly. How do you know how well your training is going? You need a way to measure the "error" between what you *want* the puppy to do (sit) and what it *actually* does. In neural networks, loss functions are that measuring stick. They quantify the difference between the network's predictions and the actual target values