IoT lab: Instead of rule based fraud detection, consider ML based anomaly detection

MSRConnections / Azure-training-course

Other

111 stars 90 forks source link

IoT lab: Instead of rule based fraud detection, consider ML based anomaly detection #95

Closed vanitech closed 8 years ago

vanitech commented 8 years ago

This may be suitable for a 300 level lab. See this blog for details http://blogs.technet.com/b/machinelearning/archive/2014/11/05/anomaly-detection-using-machine-learning-to-detect-abnormalities-in-time-series-data.aspx

jeffprosise commented 8 years ago

Fascinating! As you suggested, this would be a great scenario for a longer and more detailed 300-level lab. For now, I added a summary to the lab, mentioned how hard-coded rules could be replaced (or supplemented) by ML rules provided the anomaly detection service, and linked to the article. This should at least get students thinking about more advanced scenarios, and about how powerful it is to be able to combine Azure data services into one cohesive application.