In this workshop we will go through a use case of fraud detection in an oil distribution network. The goal is to use sensor data to identify and predict some frauds such as leakages or bypasses. You will be training models on historic data and have to deploy these models into a cloud environment to make predictions on live stream data.

This is an opportunity to familiarize yourself with end to end unsupervised Machine Learning in a production environment. You will also have a chance to productionize your trained Machine Learning model by deploying it into the cloud.

• intermediate level in ML and Data science
• working knowledge in open source Python Machine Learning stack is preferred, but R and Matlab users welcome
• no business knowledge or expertise required
• only Python can be deployed to the cloud