From the course: Data-Centric Visual AI
Join today to access over 25,500 courses taught by industry experts.
Iterative dataset improvement
From the course: Data-Centric Visual AI
Iterative dataset improvement
- [Instructor] In this module, we'll be talking about iterative dataset improvement. What happens whenever you've trained your model, but it's just not good enough? You've done the best you can to curate your dataset, you've sourced a great dataset, you've cleaned it, you've trained the model to your best of your ability, but when those results numbers come back, they're just not good enough to really push that model to production yet. Well, I'm here to let you know that it's not the end. Through iterative dataset improvement and using some of the state-of-the-art tools, you're able to unlock even more from your dataset, make that dataset even higher quality to ultimately reach that high quality model that you're chasing. To start, we're going to explore some common scenarios and how you can help different types of issues about when it comes time to train your models come in different forms. Sometimes your model's doing well in the training accuracy, but the validation accuracy's not…
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.