Fast forward: Forecasting global emerging threats
Contests
Competition between innovators is the fuel we use to foster the best innovative solutions.
The challenge
The Department of National Defence and the Canadian Armed Forces (DND/CAF) are seeking innovative forecasting models that leverage the power of publicly available information, statistical methods, and machine learning models to assist DND with forecasting flashpoints and emerging threats around the world.
What IDEaS provides
A chance to put your team’s forecasting abilities to the test for our military experts. Compete with other subject matter experts for a chance to win cash prizes. Up to 10 applicants will be selected to proceed and will receive a $25,000 grant payment to support the development of their forecasts and machine generated explanations. Following this, the best three combinations of forecasts and explanations will be awarded cash prizes for a total of:
1st place: $150,000
2nd place: $75,000
3rd place: $50,000
What innovators bring
Accurate and precise forecasting models that will improve the DND/CAF’s ability to identify emerging events and crises and provide timely information to decision-makers.
Background
The Fast forward: Forecasting global emerging threats challenge aims to prepare Canada's Defence Team, to identify and prepare for threats in an ever-changing world. The Canadian Defence Team currently uses various tools, publicly available information, and teams of analysts to provide transnational and regional intelligence, strategic warning and threat assessments including the provision of anticipatory intelligence on political conflict around the world.
Objectives
This challenge is intended to incentivize Canadian innovators to create accurate and precise forecasting models for future use by the Canadian Defence Team. The Violence and Impacts Early-Warning System (VIEWS) dataset will be used as the reference data source for all innovators to create forecasting models to predict future fatalities due to geopolitical/armed conflicts during a given time period. The variables included in the VIEWS dataset cover the following themes: past fatalities caused by political conflicts, conflict history, natural and social geography, demographics, economy, institutions, protest, and welfare.
The numerical forecasts produced for the Fast forward challenge must be accompanied and supported by machine-generated explanations, i.e., presentations in human-understandable terms of the reasoning, functioning and/or behaviour of the machine learning or statistical models used to obtain the forecasts.
The longer-term goal is to integrate these models into methods and tools aimed at helping the Canadian Defence Team intelligence analysts to perform data-driven forecasting tasks while addressing the challenges of trust and accountability for analysts leveraging these novel forms of anticipatory intelligence. The aim will be to compare and eventually adopt systems that can help analysts to rapidly organize and exploit data from publicly available and private sources.
Desired outcomes
For each forecast, the explanations must at least cover the three following aspects:
- Explanation of the forecast (e.g., in text format, graphs; the exact format(s) to be determined by each participant for their own solution).
- The intelligence value (“so what”) of the forecast.
- An explanation of what the model does (e.g., how it works, what it provides, on what the forecast is based).
The machine-generated explanations will be assessed, for the three aspects identified, using subjective metrics relative to simplicity, sensitivity, and stability as well as metrics relative to transparency, comprehensibility, and persuasiveness. This assessment will be performed by subject-matter experts (SME) from the DND/CAF intelligence analysts community.
Eligibility
The applicant must be a legal entity duly incorporated and validly existing in Canada, including:
- Canadian universities and educational institutions chartered in Canada.
- Incorporated Canadian not-for-profit organizations or associations.
- Incorporated Canadian for-profit companies, organizations, or associations.
- Provincial/territorial, or municipal government organizations. Federal and provincial crown corporations are not eligible for funding.
An eligible recipient can form a group, cluster or consortium that can include international Science & Technology collaborators.
How to apply
For full instructions on how to apply, eligibility, and all other inquiries, email the Contests team: IDEaSContests.IDEeSConcours@forces.gc.ca
Deadline
Applications must be submitted by 14:00 EST December 16, 2024.
Enquiries about the Contest can be sent to: IDEaSContests.IDEeSConcours@forces.gc.ca.
Information session: VIEWS open dataset
The challenge will leverage the open dataset provided by the Violence and Impacts Early-Warning System (VIEWS) academic research consortium between Uppsala University (Sweden) and the Peace Research Institute Oslo (Norway). A short information session will be hosted by the consortium to answer questions about the open dataset.
Date: 11:00 EST November 15, 2024
Location: Online details will be sent to those who have requested an application package.
Page details
- Date modified: