The purpose of this article is to explain how to train a model. 


In this article, we will go through an example workflow to train a multivariate model and save it to the Petro.ai database. We will then load the model back into the Petro.ai Panel, make some changes to the model parameters, and retrain the model.


Requirements 

  • RAI Product Version - 3.1
  • PetroDatabase - Schema Version: 3.0


Steps 


  1. Configure the Database connection in the Petro Panel.
    (Go to the Petro Panel Icon > Database > Use the dropdown to configure a database)
  2. Create a new model.
    (Go to the Models(beta) tab > under the new model, create a model name)
  3. Configure the new model options.
    (In the new tab Edit: <Model name>, Input your model option, Model Inputs, and Model Outputs)
  4. Save the model to the database.
    (After configuring your model, click Save)
  5. Train your model from the database.
    (In the same menu, click Train. On the bottom of the Petro Panel, click yes to confirm)


  6. Confirm that the model was trained in the Database.
    (Go to the Loaded Models > Check in the model: Train = Green Check Mark)
  7. View Results (Column) from your data table visualization.


Petro.ai Enabled

Continue through the steps if you have Petro.ai enabled within your company. 


  1. Go to your Petro.ai Suite (Your company location of where Petro.ai is installed).
  2. Search for your model within the Petro Database.
    (Go to Machine Learning > Models > Search for your model)
  3. Check your models diagnostics and view results.
    (In the same menu, click View)

For additional information on Petro.ai Panel documentation, click here.