Back

Optimizing the number of models included in outbreak forecasting ensembles

Fox, S. J.; Kim, M.; Meyers, L. A.; Reich, N. G.; Ray, E. L.

2024-01-07 epidemiology
10.1101/2024.01.05.24300909 medRxiv
Show abstract

Based on historical influenza and COVID-19 forecasts, we quantify the relationship between the number of models in an ensemble and its accuracy and introduce an ensemble approach that can outperform the current standard. Our results can assist collaborative forecasting efforts by identifying target participation rates and improving ensemble forecast performance.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.