From the course: Generative AI vs. Traditional AI
Join today to access over 24,900 courses taught by industry experts.
Self-supervised learning
From the course: Generative AI vs. Traditional AI
Self-supervised learning
- Earlier you saw that you could use machine learning to do a form of supervised or unsupervised learning. Remember that you were using machine learning to try and classify your stamps or cluster them based on similar traits. Then you saw with generative adversarial networks, or GANs, you could mix and match these two techniques. Remember that you could use a little bit of supervised learning to train the discriminator on how to make an educated guess. Now, newer generative AI systems take this a step further. The main challenge is that there's a lot more unlabeled data in the world, so even though there might be millions of of images of cats online, only a small percentage of them are labeled with the word cat. And remember that supervised machine learning systems need to be trained with labeled data. Imagine if you could use unsupervised learning to label all that unlabeled data. This is what you call self-supervised learning. So think about if you wanted to use unsupervised…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
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.