Set up Azure Machine Learning Service for Dynamics 365 Supply Chain Demand Forecast

Dynamics 365 ERP - Development

Azure Machine Learning Studio (classic) has been deprecated and replaced by the Azure Machine Learning Service, offering similar functionality and more.
To set up the Azure Machine Learning Service for Supply Chain Management forecasting, all prerequisites must be installed, including using Visual Studio Code as a code editor to modify Python files.
Cloning the GitHub repository for Dynamics 365 Supply Chain Management Demand Forecasting templates is essential for the installation process.
Creating a virtual environment in Python and installing necessary extensions within the cloned repository folder is required before executing the quicksetup.ps1 file.
Errors during setup, such as naming conflicts or soft-deleted workspaces, can be resolved by editing scripts or permanently deleting the affected workspace in Azure Machine Learning.
After resolving setup errors, the deployment can be successfully completed, and the Demand forecast parameters provided at the end of the script execution need to be configured in Azure Machine Learning studio.
Uploading files, running scripts, and setting parameters in Azure Machine Learning Studio and Supply Chain Management are crucial steps to connect the machine learning service and generate demand forecasts.
Finally, generating a statistical baseline forecast in Supply Chain Management can be done following specific procedures outlined in the article.

Registered users can view the full text for FREE!