by: Kevin Broløs & Chris Cave
(Feyn version 2.0 or newer)
QLattice can be considered as a probability distribution of
Model structures, that can be sampled from with the
sample_model function. A structure is defined as the input names, function names and their combinations inside a
Model. This excludes its weights and biases.
When you update the
QLattice with a list of
Models, the probability density function is updated to encourage those structures. Without updating, the QLattice will keep generating models with a random structure.
import feyn from feyn.datasets import make_classification train, test = make_classification() ql = feyn.connect_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(models, True, True) models = feyn.best_diverse_models(models, 10, None) ql.update( models=models )
The list of
Models to update the