Proposes Latent Interacting Particle Systems with an efficient parameterization of twist potentials to enable approximate posterior inference for coupled continuous-time hidden Markov models via twisted sequential Monte Carlo, demonstrated on a latent SIRS graph model and real wildfire data.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2025 2representative citing papers
A deep learning framework forecasts final wildfire burned area extent from ignition-time data, with an ablation showing that a four-day pre- to five-day post-ignition temporal window improves F1 and IoU by nearly 5% over a single-day baseline on held-out Mediterranean test data.
citing papers explorer
-
Efficient Inference for Coupled Hidden Markov Models in Continuous Time and Discrete Space
Proposes Latent Interacting Particle Systems with an efficient parameterization of twist potentials to enable approximate posterior inference for coupled continuous-time hidden Markov models via twisted sequential Monte Carlo, demonstrated on a latent SIRS graph model and real wildfire data.
-
Wildfire spread forecasting with Deep Learning
A deep learning framework forecasts final wildfire burned area extent from ignition-time data, with an ablation showing that a four-day pre- to five-day post-ignition temporal window improves F1 and IoU by nearly 5% over a single-day baseline on held-out Mediterranean test data.