Retraining a Model

[#1]

This article describes how to retrain a model after evaluating the model and determining changes that are required to improve predictions.

In this article:

• Determine Needed Changes and Prepare for Retraining
• Retrain the Model
• View Retrained Results for Deployment Readiness

Determine Needed Changes and Prepare for Retraining

Use the following sections as a guide to prepare your model for retraining. To edit a trained model or view included and excluded attributes, see Review and Start Training.

Retrain the Model

There are three ways to retrain your model:

1. From the Model Dashboard, click Retrain.
   
2. If a training fails, you can retrain a model from the Model Details screen.
   
   
3. From the Training Details screen, after reviewing the training details you can choose to retrain.
   

After you select to retrain your model, complete the following steps:

1. Adjust training options if needed, such as:
   • Training Date Range
   • Attributes to Exclude
2. Click Retrain. Your model is now Version 2 with a status of Requires Publish.
3. Save and Publish your changes.
   Your model is now Version 2 with a status of In Training".

View Retrained Results for Deployment Readiness

After your new version is trained, review the results of retraining following the guildelines in Evaluating the Trained Version. Continue to refine your data and model properties if needed. 

Next Steps

When you are satisfied with your results and ready to deploy, go to Deploying a Model.