Accounting for variability in conflict dynamics: A pattern-based predictive model
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Existing models for predicting conflict fatalities frequently produce conservative forecasts that gravitate towards the mean. While these approaches have a low average prediction error, they offer limited insights into temporal variations in conflict-related fatalities. Yet, accounting for variability is particularly relevant for policymakers, providing an indication on when to intervene. In this article, we introduce a novel risk-taking methodology, the 'Shape finder', designed to capture variability in fatality data, or rather the sudden surges and declines in the number of deaths over time. The method involves isolating historically analogous sequences of fatalities to create a reference repository. Comparing the shape of the input sequence to the historical references, the most similar historical cases are selected. Predictions are then generated using the average future outcomes of the selected matches. The Shape finder is derived from the theoretical understanding that strategic and adaptive interactions between the government and a non-state armed group produce recurring temporal patterns in fatality data, which are indicative of broader developments. In this article, we demonstrate that our approach maintains high accuracy while significantly enhancing the ability to predict shifts, surges, and declines in conflict fatalities over time. We show that combining the Shape finder with existing approaches, the Violence Early-Warning System ensemble, achieves a lower mean squared error and better accounts for variability in fatality data. The Shape finder methodology performs particularly well for high intensity cases, or rather country-months with substantial armed violence.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.042 | 0.015 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it