Train Predictor
A 3-step recipe with Amazon Forecast. We can help you get access.
Recipe · 3 steps
Train Predictor
with Amazon Forecast · Artificial Intelligence Management
Create an Amazon Forecast predictor and wait until training is ACTIVE.
- Create the predictor and start training using the supplied configuration.
- List predictors and confirm the newly created predictor has reached the ACTIVE status. Repeat this step while training is still in progress.
- Apply the supplied metadata tags to the now-ACTIVE predictor.
What you'll need to have ready
- predictorName: A name for the predictor.
- forecastHorizon: The number of time-steps to predict.
- algorithmArn: The ARN of the algorithm to use for training.
- performAutoML: Whether to let Forecast select the optimal algorithm.
- inputDataConfig: The dataset group and supplementary features for training.
- featurizationConfig: The featurization configuration for the predictor.
- tags: The metadata tags to apply once the predictor is ACTIVE.
For your developer
Arazzo workflow amazon-forecast:train-predictor, 3 operations: CreatePredictor → ListPredictors → TagResource. Specs on APIs.io.