Getting started with building, deploying, and trusting models

Getting started with building, deploying, and trusting models

To get started with building, deploying, and trusting models, understand the overall workflow, choose a tutorial, and check out other learning resources for working with Watson Studio in Cloud Pak for Data.

Overview of the model workflow

The model workflow has three main steps: build a model asset, deploy the model, and build trust in the model.

Overview of model workflow

Build a model asset

  1. Create a project.
  2. Add data to the project. If necessary, prepare your data.
  3. Choose a tool to build a model. You can choose from code editors, graphical canvas, or automatic tools.

Deploy the model

  1. Create a deployment space and add the model to it.
  2. Deploy and score the model, and review prediction scores and insights.
  3. Monitor deployment jobs in a dashboard.

Build trust in your models

  1. Evaluate your deployment for bias or drift.
  2. Update your data and retrain the model until you reach your quality goals.
  3. Update deployments with better-performing models.
  4. Continue to evaluate, retrain, and update the deployed model.

Tutorials

This tutorial provides a description of the tool, a video, the instructions, and additional learning resources:

Tutorial Description Expertise for tutorial
Build and deploy a machine learning model with AutoAI Automatically build model candidates with the AutoAI tool. Build, deploy, and test a model without coding.
Build and deploy a machine learning model in a notebook Build a model by updating and running a notebook that uses Python code and the Watson Machine Learning APIs. Build, deploy, and test a scikit-learn model using Python code.
Build and deploy a machine learning model with SPSS Modeler Build a C5.0 model using the SPSS Modeler tool. Drop data and operation nodes on a canvas and select properties.
Build and deploy a Decision Optimization model Automatically build scenarios with the Modeling Assistant. Solve and explore scenarios, then deploy and test a model without coding.


Learning resources

Documentation

Videos

Training