News & Events
The latest updates from the VIEWS consortium
-
Posted on 25 Nov 2024
The Hidden Costs of Armed Conflict: A Multidisciplinary Look at War’s Impact on Human Development
A literature review by VIEWS’ Societies at Risk project emphasizes that armed conflict severely impacts human development beyond immediate violence, affecting health, education, and economic stability. It identifies nine interrelated dimensions of society harmed by war and calls for comprehensive research to understand these cascading effects better. Effective recovery strategies are essential for long-term recovery.
-
Posted on 22 Nov 2024
VIEWS Co-Organizes Conference on Harnessing Arms Flow Data for Early Warning
From November 6-8, 2024, VIEWS co-organized a conference in Geneva focused on enhancing the use of arms flow data for conflict early warning. Attended by over 50 experts, discussions highlighted the need for optimizing existing data for use in quantitative models, and emphasized the value of cooperation among academics, policymakers, and practitioners to jointly improve…
-
Posted on 1 Nov 2024
Forecasting Electoral Violence
The study by Randahl et al. develops machine-learning models to forecast electoral violence globally, addressing its threat to the legitimacy and fairness of electoral outcomes. By analyzing economic indicators, historical violence, and political instability, the models predict violence risk for 2024-2025 with high accuracy, aiding in effective prevention strategies for at-risk elections.
-
Posted on 31 Oct 2024
Launch of Prediction Challenge Hosted by the Canadian DND/CAF: “Fast Forward – Forecasting Global Emerging Threats”
The Canadian DND/CAF launches the Fast Forward challenge, inviting innovators to use public data and machine learning to forecast fatalities in future armed conflicts. Ten teams may receive $25,000 each for model development, with top three teams winning up to $150,000. Applications are due by December 2, 2024.
-
Posted on 30 Oct 2024
VIEWS at PRIO AI Days: “Trust and Transparency in AI: Aligning with Stakeholder Needs”
Recordings from the PRIO AI Days sessions are now available on PRIO’s YouTube channel. Watch VIEWS researcher Simon Polichinel von der Maase discuss what’s needed, from a model development perspective, to better align AI applications in early warning with stakeholder needs.
-
Posted on 30 Oct 2024
VIEWS at PRIO AI Days: “AI in Early Warning and Early Action: Successes, Challenges and Future Pathways”
Watch the recording of the PRIO AI Days panel on the use of AI for early warning and early action, moderated by VIEWS’ Operations & Outreach Manager, Angelica Lindqvist-McGowan.
-
Posted on 30 Oct 2024
VIEWS at PRIO AI Days Panel: “AI in Humanitarian Funding: Ensuring Fairness and Transparency”
Watch the recordings from the PRIO AI Days sessions. This panel, titled “AI in Humanitarian Funding: Ensuring Fairness and Transparency,” discusses AI’s role in enhancing humanitarian funding efficiency while addressing ethical concerns. Speakers include Dr. Paola Vesco and experts from NORAD and NYU.
-
Posted on 30 Oct 2024
VIEWS talk at “UNDP Data Thursdays”
On October 17th, VIEWS Director Håvard Hegre presented VIEWS at the weekly UNDP data talk series “Data Thursdays” as one of three speakers on “Using Data to Measure Multidimensional Risks and Forecast Scenarios”. Browse the slides from the VIEWS presentation.
-
Posted on 30 Oct 2024
“Bin-Conditional Conformal Prediction of Fatalities from Armed Conflict”
Title: “Bin-Conditional Conformal Prediction of Fatalities from Armed Conflict” Authors: David Randahl, Jonathan P. Williams, Håvard Hegre Year: 2024 Publisher: – / Currently available as a preprint. Abstract: Forecasting of armed conflicts is an important area of research that has the potential to save lives and prevent suffering. However, most existing forecasting models provide only…
-
Posted on 30 Oct 2024
“This is not normal! (Re-) Evaluating the lower n guidelines for regression analysis”
Title: “This is not normal! (Re-) Evaluating the lower n guidelines for regression analysis” Authors: David Randahl Year: 2024 Publisher: – / Currently available as a preprint. Abstract: The commonly cited rule of thumb for regression analysis, which suggests that a sample size of n≥30 is sufficient to ensure valid inferences, is frequently referenced but…
