Skip to content

Click in-app to access the full platform documentation for your version of DataRobot.

Batch Prediction module

The Batch Prediction module allows you to generate batch predictions to use as an input dataset, as well as perform pre- and post- processing steps in a pipeline. For example, if you connect the output port of a Batch Prediction module to the input port of an AI Catalog Export module, the prediction results are exported as a dataset in the AI Catalog after running the pipeline.

In the Details tab, select the deployment you want to use to generate predictions, then use the provided options, which vary for non-time series and time series deployments, to finish configuring the Batch prediction module:

Option Description
Include input features Writes input features to the prediction results output alongside predictions.
  • All features: Includes every feature from the input dataset.
  • Specific features: Includes only specified features from the input dataset. Begin typing the feature name in the field and select features from the list that appears below.
Include prediction explanations Add columns for prediction explanations to your prediction output. When enabled, you can define the number of prediction explanations generated, as well as the high and low prediction threshold.
Track data drift and accuracy for predictions Tracks drift and accuracy for predictions generated in the pipeline.
Time series
Time Series prediction method
  • Forecast point: Allows you to choose a specific date from which you want to begin making predictions.
  • Forecast range: Allows you to make bulk, historical predictions (instead of forecasting future rows from the forecast point).
Select forecast point Available if you select Forecast point as the prediction method:
  • Automatically: DataRobot selects the forecast point for you based on the scoring data.
  • Manually: Allows you to select a specific forecast point date.
Select forecast range Available if you select Forecast range as the prediction method:
  • Use all dates from the prediction source: Uses all forecast distances from the input dataset to make predictions.
  • Use specific date range: Allows you to specify a specific date range for predictions.

Updated May 20, 2022
Back to top