From the course: Neural Networks and Convolutional Neural Networks Essential Training
Join today to access over 25,300 courses taught by industry experts.
Loss functions
From the course: Neural Networks and Convolutional Neural Networks Essential Training
Loss functions
The model that we've created isn't smart. It doesn't know what sport shoes are or trainers or ankle boots, and it can't tell the difference between them. If we take an image of an ankle boot as our input, it passes through our model and we'll get a prediction. Now, what makes neural networks special is that we can train it so that whenever you provide an ankle boot as an input, it learns this. Now, the way we do this is to train it with loads of examples of ankle boot images so that it gets good at this. And as you can see, there are a couple of features with an ankle boot, for example, they end at or just above your ankle, and some of them have heels. So there's gonna be some way when whenever we pass an image of an ankle boot, it predicts ankle boots as the output. And when we pass it to an image of a sport shoe, it predicts these sport shoes and not ankle boots or a bag or a dress as the output. Now, this is known as the loss function. So a loss function takes in as input what the…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.