Summary

In this chapter, we understood some of the miscellaneous features of H2O AutoML. We started by understanding the scikit-learn library and getting an idea of its implementation. Then, we saw how we can use the H2OAutoMLClassifier library and the H2OAutoMLRegressor library in a scikit-learn implementation to train AutoML models.

Then, we explored H2O AutoML’s logging system. After that, we implemented a simple experiment where we triggered AutoML training; once it was finished, we extracted the event logs and the training logs in both the Python and R programming languages. Then, we understood the contents of those logs and how that information benefits us in keeping an eye on H2O AutoML functionality.

In the next chapter, we shall further focus on using H2O in production and how we can do so using H2O’s Model Object Optimized.