You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I think another benefit of using LSTM batches for ensemble members is that there is very little cost to adding ensemble members when running the model (and to some extent, training the model). This is opposed to process based models that can take a long time to run and may limit the number of ensembles one can run in some cases, or require parallel computing to run in a reasonable time. The EnKF was developed in part to accomodate for these large models (e.g. meteorological models) and only allow for a limited number of ensembles to run while still propagating uncertainty.
Makes me wonder if we should use a particle filter (1000's of members) rather than EnKF.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I think another benefit of using LSTM batches for ensemble members is that there is very little cost to adding ensemble members when running the model (and to some extent, training the model). This is opposed to process based models that can take a long time to run and may limit the number of ensembles one can run in some cases, or require parallel computing to run in a reasonable time. The EnKF was developed in part to accomodate for these large models (e.g. meteorological models) and only allow for a limited number of ensembles to run while still propagating uncertainty.
Makes me wonder if we should use a particle filter (1000's of members) rather than EnKF.
Beta Was this translation helpful? Give feedback.
All reactions