What is a Machine Learning Experiment?Machine learning experiment involves importing a data set, training multiple machine learning models to predict one or more labels from that data set and manually evaluating all those models until one finds the most accurate model.
H2O Driverless AI transforms features, trains different machine learning models and evaluates all those models systematically. The winning model is combined with a feature engineering pipeline in order to create a seamless scoring pipeline for production .
Similarly, H2O-3 trains different H2O-3 models and does the evaluation automatically. The main difference between H2O Driverless and H2O-3 is the automatic feature engineering .
Copyright 2018-2019 H2O.ai. All rights reserved.