This repository contains the challenges and data for the hands-on lab held at the dutch azure meetup. The goal of the workshop is to build a model to predict house prices and deploy the model to Azure.
This lab includes slides with an introduction.
To complete the challenges in this repository you need to have the following:
- A Windows or Mac computer
- An active Azure subscription
Before you start working on the challenge you need to install some software on your computer. Please follow the instructions on how to setup your environment.
After you're done with the setup, start with the first challenge.
- Load and prepare the data
- Explore the data using python notebooks
- Build a model with Python and scikit-learn
- Deploy your model to Azure
The data for this workshop comes from harlfoxem who hosted all the house sales data on Kaggle.com.
This data isn't meant for commercial use. This means that while you can use the workshop format in any way you like, I kindly remind you to get other data if you plan to sell this workshop format for money.
2018-02-22
- Fixed: ECDF plot was incorrect. As it turns out, the formula was wrong.
- Fixed: Code sample in challenge 2 to get value_count() for yr_renovated was unclear.
- Fixed: Missing instructions for installing Seaborn were added in Challenge 2.
- Fixed: Added instructions to fix a problem with a missing resource provider on Azure.
2018-02-21
- Intial version