Project Details


CIF: Medium: Collaborative Research: Quickest Change Detection Techniques with Signal Processing ApplicationsProject abstractThe problem of detecting changes in stochastic systems, often referred to as sequential change detection or quickest change detection, arises in various branches of science and engineering. In all these applications, an anomaly in the environment changes in some way the distribution of the sequentially acquired observations. The goal is to detect the change and raise an alarm as soon as possible, so that any necessary action can be taken in time, while controlling the rate of false alarms below an acceptable level. While the quickest change detection problem has been actively studied since early 1950s, there are many open challenges in this field that are of theoretical as well as practical interest. This research addresses long-standing open problems in quickest change detection, as well as challenging problems that are motivated by modern applications, such as the following: 1) Optimum quickest change detection for Markov data; 2) Optimum quickest detection for transient changes; 3) (Asymptotically) optimum quickest change detection schemes for multistream data when changes are sparse; 4) Joint quickest change detection and isolation (localization of the change) in multistream data; 5) Controlled sensing for quickest change detection and isolation with composite post change hypothesis; 6) Data-driven quickest outlier detection and isolation. Furthermore, the investigators study the applications of their results in the following areas: 1) Line outage detection in power systems; 2) Epidemic detection; 3) Change detection in financial applications; 4) Surveillance using sensor networks; 5) Dynamic spectrum sensing; 6) Intrusion detection in power grids/networks; 6) Anomaly and fraud detection in big data.
Effective start/end date8/1/157/31/19


  • National Science Foundation (National Science Foundation (NSF))

Fingerprint Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.