Skip to content
This repository has been archived by the owner on Dec 1, 2024. It is now read-only.

Latest commit

 

History

History
110 lines (87 loc) · 5.35 KB

File metadata and controls

110 lines (87 loc) · 5.35 KB

Build, Deploy, and Run a Machine Learning Model with AutoAI

Tutorial Scenario

Golden Bank is a leading mortgage provider through their network of neighborhood branches. This tutorial cover these goals:

  • The bank uses AI to process loan applications and needs to avoid unanticipated risk and ensure that its applicants are being treated fairly.
  • The bank has sample data in a file and wants to create a prediction model to identify the highest value customers to target for mortgage offerings with minimum development resources and time.

Exercise: Create AutoAI Experiment

  1. Associate machine learning service with project

    1. If you haven't created required services and the sample project, go to the previous session and complete Prepare Services and Create Project.
    2. From the Cloud Pak for Data navigation menu, choose Projects > View all projects.
    3. Open the MLOps and Trustworthy AI project.
    4. From the Manage tab, choose Services & integrations.
    5. Click Associate service.
    6. Select the machine learning service and click Associate.
  2. Create the AutoAI experiment

    1. From the Assets tab, click New asset.
    2. Click AutoAI.
    3. Enter “Mortgage Approval AutoAI” as name and click Create.
  3. Add a data source to the experiment

    1. We need to add a data source the experiment and configure AutoAI to build and train models based on the data source
    2. Click Select from project to add a data source.
    3. Click Data asset and check GoldenBank_HoldoutData.csv.
    4. Click Select asset.
  4. Verify your screen looks like the following image.

Exercise: Configure AutoAI Experiment

  1. Configure the AutoAI experiment

    1. Now we need to configure the experiment.
    2. Click No for Create a time series forecast?
    3. Choose MORTGAGE_APPROVAL for What do you want to predict?
    4. Click Experiment settings.
    5. We are doing a binary classification so leave the default prediction type checked.
    6. We use the default settings:
      • Positive class value: 1
      • Optimized metric: Accuracy
      • Optimized algorithm selection: Score and run time
    7. Scroll down to select Gradient Boosting Classifier and XGB Classifier as the algorithms for the experiment to run.
    8. Make sure 2 is highlighted and click Save settings.
  2. Verify your screen looks like the following image.

Exercise: Run AutoAI Experiment

  1. Run the AutoAI experiment

    1. Click Run Experiment.
    2. It will next take a few minutes to run through various pipelines.
    3. Click the pipelines on the Relationship map and Pipeline leaderboard to see more information.
  2. Verify your screen looks like the following image.

Exercise: Review Pipeline Details

  1. Review the pipeline details

    1. Click the top ranked pipeline, Pipeline 7, with Gradient Boosting Classifier, plus hyperparameter optimization and feature engineering.
    2. Check Model information.
    3. Check Feature summary.
    4. Check Model evaluation.
  2. Verify your screen looks like the following image.

Exercise: Save Pipeline as a Model

  1. Save the pipeline

    1. Click Save as.
    2. Take the default values and click Create.
    3. Click View in project after the model is saved successfully.
    4. Now we have a new model created by AutoAI, ready to be deployed.
  2. Verify your screen looks like the following image.

Exercise: Promote and Deploy Model

  1. Promote the model to a deployment space

    1. Click Promote to deployment space.
    2. For the Target space, select “Golden Bank Preproduction Space”, which is created in the previous session.
    3. Check the Go to model in the space after promoting it option.
    4. Click Promote.
  2. Create an online deployment for the model

    1. On the deployment space screen, click New deployment.
    2. For the Deployment type, select Online.
    3. For the Name, enter “Mortgage Approval AutoAI Deployment” with no leading or trailing spaces.
    4. For the Serving Name, enter “mortgage_approval_autoai”, append some characters to make it unique if name is taken.
    5. Click Create.
  3. Verify your screen looks like the following image.

Exercise: Run the Model

  1. Make a prediction request to the model – use json

    1. Click Mortgage Approval AutoAI Deployment.
    2. On the Test tab, click Paste JSON.
    3. Click Browse local files.
    4. Click GoldenBank_AutoAIData.json which you downloaded from the this repo and Confirm.
    5. Click Predict to make a prediction request for the entry in json.
    6. It comes back with prediction of 0 and 64% confidence.
    7. Feel free to change some values and see if prediction results would be different.
  2. Verify your screen looks like the following image.