Answers are grounded in this paper and cite their sources.
Mathematical models are used to simulate real-world systems in many scientific fields.These models can be fitted to real-time data, and used to generate probabilistic forecasts that describe how the system will behave in the future and convey the uncertainty in these predictions.Particle filters are a class of Sequential Monte Carlo (SMC) methods (Doucet et al., 2001) that have been used in many scientific fields for real-time forecasting, with the COVID-19 pandemic being one of the most recent and high-profile examples.These methods can be used to estimate model parameters and state as new data become available ("online estimation").pypfilt is a Python package for online estimation and forecasting that implements several particle filters.It was developed to enable real-time seasonal influenza forecasting in Australia, for which we won two national innovation awards (Pyne, 2018), and played a key role in generating forecasts that have supported Australia's COVID-19 response (Moss et al., 2023).The package is deliberately generic and readily applicable to other domains.
No takes yet. Share an insight, caveat, or question.
Robert Moss (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: