The biggest advantage of NAS is that one does not need to spend copious amounts of time designing a neural network for a particular problem. This also means that those who are not data scientists can also create machine learning agents as long as they can prepare data. In fact, Google has already productized this framework as Cloud AutoML, which allows anyone to train customized machine learning models with minimum effort. According to Google, Cloud AutoML provides the following benefits:
- Users only need to interact with a simple GUI to create machine learning models.
- Users can have Cloud AutoML annotate their own datasets if they are not labeled already. This is similar to Amazon's Mechanical Turk service.
- Models generated ...