How (Not) to Hybridize Neural and Mechanistic Models for Epidemiological Forecasting

TL;DR

This paper proposes a method for epidemiological forecasting that explicitly models non-stationarity by decomposing infection data into components and using them as control signals, improving long-term prediction accuracy.

Contribution

It introduces a novel hybrid modeling approach that extracts multi-scale structure from infection data to enhance neural-mechanistic epidemiological models.

Findings

Reduces long-horizon RMSE by 15-35%

Improves peak timing error by 1-3 weeks

Lowers peak magnitude bias by up to 30%

Abstract

Epidemiological forecasting from surveillance data is a hard problem and hybridizing mechanistic compartmental models with neural models is a natural direction. The mechanistic structure helps keep trajectories epidemiologically plausible, while neural components can capture non-stationary, data-adaptive effects. In practice, however, many seemingly straightforward couplings fail under partial observability and continually shifting transmission dynamics driven by behavior, waning immunity, seasonality, and interventions. We catalog these failure modes and show that robust performance requires making non-stationarity explicit: we extract multi-scale structure from the observed infection series and use it as an interpretable control signal for a controlled neural ODE coupled to an epidemiological model. Concretely, we decompose infections into trend, seasonal, and residual components and use these signals to drive continuous-time latent dynamics while jointly forecasting and inferring time-varying transmission, recovery, and immunity-loss rates. Across seasonal and non-seasonal settings, including early outbreaks and multi-wave regimes, our approach reduces long-horizon RMSE by 15-35%, improves peak timing error by 1-3 weeks, and lowers peak magnitude bias by up to 30% relative to strong time-series, neural ODE, and hybrid baselines, without relying on auxiliary covariates.