Charles River Analytics Inc., developer of intelligent systems solutions, has received funding under the Intelligence Advanced Research Projects Activity’s FOCUS program. Our Reasoning about Multiple Paths and Alternatives to Generate Effective Forecasts (RAMPAGE) process can help analysts empirically evaluate approaches to counterfactual forecasting and lessons learned. We are partnered with the University of Maryland, Georgia Tech, Michigan Technological University, and Georgia Tech Research Institute on the RAMPAGE effort, which is valued at $4.5M with an option exercised.
“Common human biases, such as confirmation bias and anchoring, limit the reasoning paths analysts follow,” said Dr. Ashley McDermott, Scientist at Charles River Analytics. “Worse yet, biases can result in single-path reasoning, where analysts focus on a limited flow of events defined by previous outcomes and observations. Counterfactual forecasting predicts what might have happened under different circumstances or what might have been predicted using different methods. This forecasting method provides a foundation to help analysts understand the world and its complex interconnections, as well as lessons learned—implications to improve future analyses based on this deep understanding.”
“Counterfactual forecasting and lessons learned are critical tasks, yet little research investigates effective methods and processes for performing them,” continued Dr. McDermott. “Our RAMPAGE approach draws on our team’s empirically driven model of hypothesis generation to help analysts generate alternate outcomes.”
We are leveraging our rich cognitive modeling expertise to rigorously develop and test methods that make it easier for analysts to achieve multi-path reasoning for RAMPAGE.
Contact us to learn more about RAMPAGE and our other Cognitive Modeling capabilities.
Supported by the Intelligence Advanced Research Projects Activity (IARPA) via Department of Interior/ Interior Business Center (DOI/IBC) contract number 140d0419C0048. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of IARPA, DOI/IBC, or the U.S. Government.
