0

Is it possible (and how?) to provide time series for binary classification in H2O.ai's Driverless AI? I have dataframe that looks like this:

  • ID
  • Status/Target [0/1]
  • TimeStamp for events that happened on given ID, in last 90 days
  • Details of those events (category, description, values, etc...)

Ideally what i want is to build a model that predict status for given ID, based on provided history of events.

Erin LeDell
  • 8,704
  • 1
  • 19
  • 35
Ingelik
  • 1

1 Answers1

2

For H2O's Driverless AI, you can use it out of the box for time-series modeling. See this section. You need to provide the "Time Column" as your TimeStamp and add ID to your "Time Groups Column".

If your target column is 0s or 1s, then it should automatically identify it as binary. If you not, you can toggle it from regression to binary classification.

Neema Mashayekhi
  • 902
  • 4
  • 5
  • It looks like with those settings, i receive individual timeseries prediction per ID (each ID has diff. proba for given TimeStamp). I'm looking more for something analogous as if I use featuretools to derive time dependant features, and those would be incorporated into model. – Ingelik Mar 30 '21 at 12:43
  • @Ingelik, I am not sure I understand what you are trying to do. DAI creates the transformed features and also models which it incorporates the new features. Another option is create the model/experiment, then use "Transform on another dataset" http://docs.h2o.ai/driverless-ai/latest-stable/docs/userguide/transform-another-dataset.html, but this can lead to overfitting if you reuse these for a new model. If you are looking for only feature engineering (no modeling), I don't think there is an option for that at the moment. – Neema Mashayekhi Mar 31 '21 at 16:52