This section lists all the AutoML systems considered for the present survey and the associated metadata.
A description of the characteristics considered for each system can be found in the Comparison section.
Help
We need your help!
If you want to contribute to this section, you can either edit any of the systems' characteristics by clicking on each section's badge, or add new systems in our repository.
AutoKeras focuses on neural architecture search, using bayesian optimization with a custom acquistion function based on network morphism. Presents a high-level API inspired by scikit-learn, where users only need to select a task-specific model; and a low-level API where users can completely customize a neural search space.
