Capstone Projects

Predicting Network Failures Before They Happen

Program: Data Science Master's
Host Company: TDS Telecom
Location: Madison, Wisconsin (onsite)

Consistent, dependable high-speed data service is our hallmark. Failure is not only expensive, but also a PR nightmare. Our data networks can be subject to service disruptions ranging from a one-household outage to taking down an entire metropolitan area. We’ve honed our processes for reacting swiftly when troubles occur, but now we need to make a paradigm shift. You will be developing predictive analytics on a rich set of network component and customer equipment behaviors. A successful solution will enable us to predict what is going to happen, with enough lead time so the problem can be fixed before our customers are aware.