Interactive analytics for smart cities infrastructures

摘要: Infrastructures in smart cities are complex systems, more than the sum of their parts, and operate through a multitude individual collective decisions. These systems experiencing gradual but substantial change moving away from being non-interactive manually-controlled to utilizing tight integration both cyber (computation, communications, control) physical representations guided by first principles, at all scales levels. As result, facility managers for such demanding control over system operation. They prefer data-driven insights into performance usage across portfolios make effective Such help determine, if is operating efficiently enable facilities manager/system operator investigate areas improvement evaluate upgrades. In this paper we describe one system, which uses control-oriented models build data intelligence layer underlying infrastructure. Our provides interactive analytics answering queries making recommendations about We extend our previous work with using regression trees ensembles predictive modeling these large 'messy' show how tree based can be converted knowledge discovery database. present preliminary results office building.