Welcome to the workshop on using the Microsoft Azure ML Forecasting Toolkit.
In this one-day hands-on workshop, we learn how to use the Azure ML Python Forecasting library to load and pre-process time series data, train and compare multiple time series models, and deploy a forecasting solution using Azure Machine Learning.
The two pre-requisites for the workshop are Python proficiency and familiarity with Azure Machine Learning for end-to-end deployments.
- We will be using Python for data science, and in particular the
- We will use Azure Machine Learning (and its Python SDK) and its architecture for model deployment and operationalization. The LearnAI Bootcamp series is a great resource for learning about Azure ML.
Azure ML Package for Forecasting
- Ingest a time series data and run various summaries to explore and understand its shortcomings
- Learn about different options for imputing missing values for times series data
- How to do feature engineering in the context of times series data
- How to train multiple time series models and compare their performance
- How to deploy trained models using Python’s Azure Machine Learning API
Please note: This is a rough agenda, and the schedule is subject to change pending class activities and interaction.
- 9:00 - 10:00 : Introduction and Context
- 10:00 - 11:00 : Setup and data ingestion
- 11:00 - 12:00 : Data exploration
- 12:00 - 1:00 : Lunch
- 1:00 - 2:00 : Data pre-processing and feature engineering
- 2:00 - 3:00 : Training and testing time series models
- 3:00 - 4:00 : Model deployment
- 4:00 - 5:00 : Q&A and Feedback
Running the Notebook
In order to run the Notebook, follow the installation steps outlined here.