We are deeply grateful to the National Science Foundation, University of Pittsburgh Medical Center, Disruptive Health Technology Institute, Metro 21 Initiative, Richard King Mellon Foundation, and MacArthur Foundation, who have provided funding for the following projects:


Richard King Mellon Foundation, Metro 21: Knowledge-Powered Pittsburgh to Improve Urban Quality of Life, 1/1/2016-12/31/2018. Our project, “Urban Predictive Analytics for a Safer and Cleaner Pittsburgh”, will develop and deploy predictive analytics for violence prevention in Pittsburgh. By incorporating many city and county data sources, and by integrating predictive analyses at the geographic, subpopulation, and individual levels, we will provide the Pittsburgh police and the city and county leadership with situational awareness of the many inter-related factors influencing patterns of violence, assisting the development of long-term violence prevention strategies and tactical interventions. Total award: $600,000. Project award: $250,538.

UPMC Healthcare Innovation Grant, Anomalous Pattern Detection from Healthcare Data Streams, unrestricted gift awarded 11/8/2010, funded by the University of Pittsburgh Medical Center- Technology Development Center. This project will apply novel pattern detection methods to detect anomalous patterns of patient care. Total award: $121,503.

Disruptive Health Technology Institute, Discovering Anomalous Patterns of Care to Improve Health Outcomes and Reduce Costs, unrestricted gift awarded 7/7/2014. We plan to create a widely applicable methodological and implementation framework for using massive quantities of health insurance claims data to discover patterns of care with significant potential impacts on patient outcomes and healthcare costs. Total award: $20,000.


NSF IIS-0953330, CAREER: Machine Learning and Event Detection for the Public Good, 7/1/2010-6/30/2016, funded by the National Science Foundation. This project will create and explore novel methods for detection of emerging events in massive, complex, real-world datasets. This research will be integrated with a multi-pronged educational initiative to incorporate machine learning into the public policy curriculum. Total award: $529,962. (summary) (NSF page) (project page).

NSF IIS-0916345, Fast Subset Scan for Anomalous Pattern Detection, 8/1/2009-7/31/2013, funded by the National Science Foundation. This project will develop new, general subset scan methods for efficient pattern detection in massive datasets. Total award: $499,991. (summary) (NSF page) (project page).

NSF IIS-0911032, Discovering Complex Anomalous Patterns, 9/1/2009-8/31/2014, funded by the National Science Foundation. This project will develop an integrated probabilistic framework for pattern discovery, incorporating detection, characterization, explanation, and learning from user feedback. Total award: $2,598,153. (summary) (NSF page) (project page).

Metro21 Initiative, Open-Source 311 Predictive Analytics for the City of Pittsburgh, 1/1/2015-12/31/2016. We propose to deploy, evaluate, and extend an open-source version of our CityScan predictive analytics software as part of the City of Pittsburgh’s 311 call system, in collaboration with the City leadership. By predicting emerging clusters of 311 calls (non-emergency service requests) and providing support for the City’s operational decisions based on the predicted clusters, we will enable the City to respond proactively and effectively to emerging challenges and citizen needs. Total award: $41,406.

John D. and Catherine T. MacArthur Foundation, Evaluating Machine Learning Methods and Tools for Use in Human Rights Work (with CMU’s Center for Human Rights Science), 3/1/2013-8/31/2014. We will develop new machine learning methods for early detection and advance prediction of conflict events, and evaluate the potential utility of these methods for enabling proactive responses to outbreaks of violence and human rights abuses. Total award: $175,000.