Series Prediction

Description

This step provides a way to get a prediction from a Series Model for an entire dataset with one single step at once. The prediction results will be added as new columns to the dataset.

Application

Once this step is applied, the data set produced can be used in multiple different ways:

• Model Evaluation: If the ground truth results are also available in the dataset, then the resulting dataset can be used to compare prediction and ground truth. This can help evaluating and comparing models to assess if they can be used in production.

• Model Prediction: If the ground truth results are not known, the dataset produced can be used to explore new predictions. For example, this could help identifying if a test would fail before running it. The dataset produced can be used with the exploration tools of the platform, such as Line Plot, 2D Point Plot etc.

How to use

To use this step, you need at least one Series Model available in the notebook.

• Series Data: Select the data for which you want to make prediction. This data should have the same format as the data used to train the model (same columns available, and tests should have the same length, i.e. same number of rows) as in the training set. The data can contain multiple tests.

• Series Model: Select the trained series model that you would like to use to make predictions. You can only select one model. If you want to make predictions from multiple models, you can repeat this step and then join the different predicted data sets.

• Prediction Dataset Name: Write the name you want the dataset with predictions to have.