4

I have two questions on h2o.automl and I couldn’t find any documentation on these topics.

  1. I can save/reuse the leader (automl) model in R using h2o.saveModel and h2o.loadModel. But how do I save/reuse other automl models, say the 6th model in the leaderboard?

  2. I can build on top of the old model by keeping the same project_name in the syntax. How can I do the same after loading a pre-built model?

qed
  • 141
  • 1
  • 4

1 Answers1

3

If you autoML object called aml, then aml@leaderboard$model_id has the list of all model ids that were built. You can then pass each of those to h2o.getModel() (which you can then call h2o.saveModel() on).

You can see an example in the h2o tutorials.

As for the second part of your question, by "model" you mean expanding the autoML, so that more models are tried? As far as I can tell, this is not possible. You'd need to run it again, perhaps with a different explicit seed, then manually merge the leaderboards.

(If you meant improving the leader model, then it depends on the type of model. E.g. for deep learning you give it more epochs, if for gbm or random forest you give it more trees.)

Darren Cook
  • 1,772
  • 1
  • 12
  • 26