From the course: Deep Learning and Computer Vision: Object Detection with PyTorch
Join today to access over 24,900 courses taught by industry experts.
Overview of deep learning, part 2 - PyTorch Tutorial
From the course: Deep Learning and Computer Vision: Object Detection with PyTorch
Overview of deep learning, part 2
- [Instructor] Welcome back. In the previous video, we covered the basic structure of a neural network. We learned that neural networks consist of layers of neurons that work together to make required predictions. In this video, we'll be exploring how these networks learn from data through a process called back propagation. So how do neural networks learn? The answer lies in the technique called backpropagation. In simple terms, backpropagation is the process where the network learns from its past mistakes. Here's how it works. Let's say a network makes a prediction, but it's wrong. We'll calculate how far off the prediction was from the actual result. This is called the error, but instead of just stopping there, the network will take this error and propagate it backward through the layers. This allows the network to adjust the weights of the neurons in each layer so that the next time, it can make a better prediction. For example, imagine a network is trying to classify an image as a…
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.