How do you build and deploy predictive analytics into your ongoing business processes so the results can be used in real-time to improve operations? This is encountered in many business environments, ranging from machine-to-machine and Internet of Things scenarios to applications in oil & gas and utilities industries. Learn how to leverage your data assets – including the massive “data exhaust” collecting at many businesses – to build and operationalize predictive models that can improve your business operations. Learn the general challenges to build and deploy a predictive model through machine leaning and survey the range of applications we have encountered which would benefit from this capability.
Throughout this session, we will share real-world examples from a manufacturing company that created a hybrid solution using cloud computing and on-premise resources for predictive maintenance of equipment to identify impending equipment failures and avoid both unplanned downtime and unnecessary maintenance. You will learn best practices from our experience leveraging cloud computing to carry out machine learning at scale over sensor data. You will see how we deploy this predictive model from the cloud into a stream processing system that operates on premise in the manufacturing plant, processing data feeds from instruments across the facility to perform continuous failure prediction. Finally, we share design considerations we implemented to retrain and continuously improve the predictive model over time.
Architect in the eXtreme Computing Group of Microsoft Research developing tools and cloud services for the exploration of data.
For information on exhibition and sponsorship opportunities, contact Susan Stewart at firstname.lastname@example.org.
For information on trade opportunities contact Kathy Yu at mediapartners
For media-related inquiries, contact Maureen Jennings at email@example.com
View a complete list of Strata contacts.