Research at PaCE

We build transparent, data‑driven methods to forecast conflict risk. Explore our publications and methods.

Methods at a glance

We combine transparent statistical models with event and geospatial data to forecast conflict fatalities over 1–6 month horizons.Read more

Featured publication

Investigating Batch Inference in a Sequential Monte Carlo Framework for Neural Networks

Millard A, Murphy J, Green PL & Maskell S — ICASSP 2026 - IEEE International Conference on Acoustics, Speech and Signal Processing, 2026
Highlights
  • Peer‑reviewed venue
  • Recent (2026)
  • See details on Publications
Featured publication

Utilising Gradient-Based Proposals within Sequential Monte Carlo Samplers for Training of Partial Bayesian Neural Networks

Millard A, Murphy J, Zhao Z & Maskell S — ICASSP 2026 - IEEE International Conference on Acoustics, Speech and Signal Processing, 2026
Highlights
  • Peer‑reviewed venue
  • Recent (2026)
  • See details on Publications

Impact & updates

  1. 2026-01
    HICSS presentation on Structured Pixels (GeoAI causal methods)
  2. 2025-10
    Forecasts refresh with improved uncertainty calibration
  3. 2025-07
    DTW evaluation study on conflict trajectories