Extract h2o models and evaluate performance

In this final exercise, you will extract the best model from the AutoML leaderboard. The h2o library and test data has been loaded and the following code has been run:

automl_model <- h2o.automl(x = x, 
                           y = y,
                           training_frame = seeds_data_hf,
                           nfolds = 3,
                           max_runtime_secs = 60,
                           sort_metric = "mean_per_class_error",
                           seed = 42)

Extract the leaderboard from the AutoML output. Note, that you can extract subsets of objects not only with $ but also with @.

Exercise

Extract h2o models and evaluate performance

Instructions 1/4