“Making Predictions is hard, especially about the future (Yogi Berra).” There is a significant buzz about the promise “Predictive Analytics” and “Data Driven Decisions” have for making these predictions easy. However, many national-security related events are unique and unresponsive to highly automated tools and techniques. This presentation provides an overview of research conducted using IARPA’s Global Forecasting Challenge 2 competition as our laboratory. The hypotheses and supporting techniques described range from effectively integrating “the wisdom of the crowd” and “Super Forecasters” to using machine learning techniques to obtain the best forecast possible with available information.
Dr. Daniel Maxwell is the president of KaDSci, LLC, a veteran owned decision sciences research and analysis boutique. Over a thirty-year research career, Dan has completed a significant body of work and published in the areas of military operations research, simulation, military transformation, C4ISR analyses, command and control, and decision-making. Dr. Maxwell has served on the US Defense Science Board Task Force on Intelligence, the Board of Directors of the Military Operations Research Society, and multiple research companies. He maintains affiliate faculty status at George Mason University, occasionally teaching graduate courses in Operations Research and Command and Control.
Dr. Anamaria Berea has a dual PhD in economics (2010) and computational social science (2012). Her research focuses on the emergence of communication in biological and social networks, by applying theories and methods from economics, complex systems and information theory to reinterpret historical, anthropological and biological evidence regarding fundamental aspects of communication. Dr. Berea was one of the data scientists with the NASA/SETI Frontier Development Lab (2017) and an AI mentor for the astrobiology group at NASA FDL (2018). She is the author of the book “Emergence of Communication in Socio-Biological Networks,” Springer, 2018 and has experience solving data science problems for the FAA. She is currently an Associate Term Professor in the Computational and Data Sciences Department here at Mason.
Dr. Chris Karvetski is an experienced decision/data scientist with broad expertise in applied mathematics and analytics, including expert judgment and probabilistic modeling, optimization methods, entity matching, decision support methods, and experimental design. He has domain knowledge within intelligence analysis, financial crime, geopolitical forecasting, and schedule/process optimization. He holds a PhD degree in systems engineering from the University of Virginia, and a MS degree in applied mathematics from Clemson University.