"ML in a Minute" is our conversational series on answering machine learning questions. Have questions you want answered? Tweet at us.
What is AutoML (in 60 Seconds or Fewer)?
AutoML, "automated machine learning," is a machine learning tactic for building models without needing to be worried about specific implementation considerations like hyper parameters, individual model selection, and a lot of smaller details that might bog you down. Instead, autoML says, "Let's take a pragmatic approach to getting a model that you can use in production." That model could be for natural language processing, it could be computer vision, it could be any sort of deep learning framework. AutoML automates the hard parts of identification of what model you should use, and optimizes that model performance against the data that you provided.
How Can I Get Started? Does Roboflow Help with AutoML?
If you're doing autoML in computer vision, Roboflow is one of the best ways to get going. If you show up with a video – or set of images – you can leave with a model that you can use by API or even on-device on the edge. So, automated machine learning is something that we make quick and easy and even one click to train your auto ML model.
Liked this? Be sure to also check out the computer vision glossary.