Monitor your machine learning models using Watson OpenScale in IBM Cloud Pak for Data
Build an app to monitor and deploy open source machine learning models
This code pattern is part of a series that explores telecom call-drop predictions using IBM Cloud Pak for Data, data virtualization, Watson OpenScale, and Cognos Analyics.
Businesses today are increasingly certain that AI will be a driving force in the evolution of their industries over the next few years. To successfully infuse AI into your product or solution, there are many factors that challenge its widespread adoption in the business–and to achieving your expected outcomes. These may include:
Building trust – Organizations and businesses tend to be skeptical about AI because of its “black box” nature, resulting in many promising models not going into production.
Algorithm bias – Another inherent problem with AI systems is they’re only as good–or as bad–as the data they’re trained on. If the input data is filled with racial, gender, communal or ethnic biases, your model’s accuracy is going to eventually drift away.
Making decisions explainable – How can the model prove the reasoning behind its decision-making? It is critical that AI outcomes are fully explainable by keeping a complete track of the inputs and outputs of any AI-powered application.
In this code pattern, you’ll learn how to monitor your AI models in an application using Watson OpenScale in IBM Cloud Pak for Data. We’ll use an example of a Telecomm call drop prediction model. After the user has completed the code pattern, they’ll understand how to:
- Store custom models using open source technology on Watson Machine Learning
- Deploy a model and connect the model deployment to Watson OpenScale on IBM Cloud Pak for Data and on IBM Cloud
- Setup model fairness and model quality monitors with Watson OpenScale on IBM Cloud Pak for Data and on IBM Cloud, using a python notebook
- Create a project and setup a Python notebook on IBM Cloud Pak for Data
- Data stored into Cloud Pak for Data internal database.
- The joined data is stored back to the internal database of IBM Cloud Pak for Data and assigned to the current working project.
- Create machine learning models using Jupyter Python notebooks to predict call drop, one cell tower at a time.
- Model trained and/or stored in Watson Machine Learning, which is also connected to Watson OpenScale.
- Configure fairness, quality and explainability Monitors for each cell tower’s model, and present within IBM Cloud Pak for Data, or on other external Clouds (Multi-Cloud Architecture).
Find the detailed steps for this pattern in the README. The steps show you how to:
- Create a Watson Machine Learning instance.
- Create a new project in IBM Cloud Pak for Data.
- Upload the dataset to IBM Cloud Pak for Data.
- Import notebook to IBM Cloud Pak for Data.
- Follow the steps in the notebook.
- Display deployment in Watson OpenScale.
- View additional use-case for Watson OpenScale.