2016 Hello Cloud Machine Learning Award
Competition year
Active Dates (GMT):
Online Competition
Win one of two trips to Imagine Cup World Finals and to meet the Azure team!
Take what you learn from the new Hello Cloud Azure Machine Learning activities, expand on them, and submit to the Azure Machine Learning Award for a chance to win a trip to the Imagine Cup World Finals.
Winners will be chosen by the Azure Machine Learning team. The most creative and inventive web or mobile game app developed with the churn analytics AML API will win. Entries will also be evaluated on how well students use the SVM or boosted decision tree models from the tutorial.
Modify your results from the tutorial below, or create something new leveraging the APIs—just make sure to think outside the box! To participate, all you have to do is copy the URL of your final project into the submission field for the Hello Cloud Machine Learning Award. The two winning entries win an all-expenses paid trip to the Seattle area for the Imagine Cup World Finals and to meet the Azure Machine Learning team!
Azure Machine Learning Churn Prevention and Intervention
Machine Learning Churn Prevention and Intervention are two new Hello Cloud student activities where you combine introductory machine learning with a basic online game. Follow the tutorial available in the lab PDF. The first four steps involve Azure Machine Learning, and the final two steps involve publishing the web app to Azure App Service. You will learn some basics about machine learning and can access links to additional content in the lab if you want to dig deeper.
What is this project about?
“Churn” is a term used when a consumer stops using a good or service. It is a common scenario with customer data, and also takes place in online games. A gamer at risk of churning is tempted to stop playing. In order to keep a gamer involved, developers use machine learning to identify potential gamers who may churn, and then take steps in the game to intervene. Based on past player history or data for similar players over time, is a gamer likely to quit? If so, developers can automate the machine to offer them a bonus, achievement, or another incentive to stay and play.
What technologies will I use?
This contest integrates Azure Machine Learning with a web app hosted in Azure App Service. You build, train, score, and evaluate models in the Azure ML Studio, then publish a machine learning model as a web service. You then take a basic game, connect it to your Azure ML service, and publish it to the web. No prior experience with machine learning or ASP.NET web development is required.
To complete this activity, you need a Microsoft Account, an Azure Subscription, and Visual Studio 2015