by: Chris Cave & Jaan Kasak
(Feyn version 3.0 or newer)
feyn.prune_models removes redundant and poor performing
Models from a list. It expects the list to be sorted by some metric. The best
Model in the list is never removed.
Models get removed based on a decay function. A
Model decays as a function of how many epochs it has been fitted and its performance compared to its peers. It will also remove models that are structurally identical from the list, leaving only the best performing one.
This prevents us fitting
Models that are not going to improve any further.
This is typically used after fitting and prior to updating. This ensures we keep
Models with high potential, conserve computational resources, as well breaking out of local minima.
Continuing from the previous sections, we now add the
prune_models function to our workflow.
from feyn.datasets import make_classification
train, test = make_classification()
ql = feyn.QLattice()
models = ql.sample_models(train.columns, 'y', 'classification', max_complexity=10)
models = feyn.fit_models(models, train, 'binary_cross_entropy', 'bic', 4)
models = feyn.prune_models(
This is the list of
Models you want to remove the redundant and poor performing ones from.
At most this many models will be returned. If
keep_n is None, models are left to be pruned only by decay.