Create and Test Predictions

The Predictions section of the Workbench is where you deploy, manage, and monitor machine learning models to generate real-time predictions, providing actionable insights and driving data-driven decisions. The next step is to begin making the models for your project, using all the configurations you have set up in the previous steps.

In the Recommenders section of the Workbench, you will find Predictions. This is where you will manage the models for your recommender.

You will see a list of all of previously configured models, as well as being able to create new models. Creating models using this interface is more suited to cases which use a small number of models. Cases where tens or hundreds of models will need to be trained are better handled in our Jupyter Notebooks.

Create Models

To create a new model for your recommender project, select + Create Model. The details you input here will be used to generate a number of models that can be tested and selected for your project.

You will need to provide a unique Predict ID and Description. You will refer back to this Unique ID at any point in the process. You can also link the model to a previously created project in Allocated to projects.

• Specify the Model ID associated with this version of the model, and remember to add the version number to the ID.

• You will almost always have multiple versions. Assigning a Version will ensure you can appropriately track changes, as you experiment with the parameters of your models.

• Choose a Model Type from the list of supported model and provide the Model Category. These details will allow you to more effectively organize your overall model training.

• Select the Primary Data Frame .hex file that you created in the Feature Store and update the Version number.

Generate & Specify Model Parameters

In Model Parameters you can specify the details and add notes to be stored with the model. Use the Describe model purpose, parameters and other actions to take notes.

For the Model Parameters functionality, you can click Generate Default to generate a list of the model parameters and their default values. These can be modified to change the behaviour of the model. Examples of model parameter set ups can be found in the other pre-configured example projects.

Taking Notes & Retrieving Features

Notes should include the features you will be using in the model. All of which can be retrieved using the Retrieve Features button. Notes should also include the model parameters, as well as a summary of the changes made, in each version of the model.

Save your model configuration in the top right corner.

Generate Models

Begin building out your model and view the Result, by clicking the Generate Model button. This will start the model training process, which may take some time to complete.

To view training progress, and any interim models being produced, continue to click the Result button. This will update the list and present the models being trained. Once the models have been trained they will appear in the table.

Clicking on a model will show you the metrics and its effectiveness. This action is how you can select which of the models is the preferred one for your needs.

Deploy Models

Once a model has been selected, Save and Deploy Model, to have it ready to use in the next (deployment) step. You can view all deployed models in the window below your trained models list.