In this article we will see:
- how to deploy a simple classification model on Azure Machine Learning Studio using PyCaret.
The model will be consumed in a Power BI report to predict new data.
- how to implement automated machine learning with Azure AutoML feature and how to use it in Power BI report.
- how to create an AutoML model with PowerBI Service dataflow and how to use it in Power BI report.
The dataset used in this demonstration is the Diabetes Health Indicators Dataset.
The dataset was randomly split into:
- 60% for model training (about 152k rows)
- 2% as hold-out sample (about 5k rows)
- 38% as unseen data (about 96k unlabeled rows)
Working with PyCaret in Azure ML Studio
This is a step by step tutorial on how write, train and deploy ML model using PyCaret in Azure Machine Learning Studio
How to use in Power BI desktop:
You can download pbix file from here and see python visual script to plot ml metrics
Workign with Azure AutoML feature
You can download pbix file from here and see python visual script to plot ml metrics.
Workign with Dataflow
Explore Prediction report (Open Report in full screen mode):
You can download pbix file from here