Class to stop training on a network when the validation error increases.
Normally, during training of a network, the error will always decrease
on the set of data used in the training. However, if an independent
set of data is used for validation, the error will decrease to a point,
and then start to increase. This increase normally occurs due to the
fact that the network is starting to learn noise in the training data
set. This stopping criterion function will stop when the validation
error increases.
|
Methods
|
|
__init__
stopping_criteria
|
|
|
__init__
|
__init__ (
self,
max_iterations=None,
min_iterations=0,
verbose=0,
)
Initialize the stopping criterion class.
Arguments:
max_iterations - The maximum number of iterations that
should be performed, regardless of error. min_iterations - The minimum number of iterations to perform,
to prevent premature stoppage of training. verbose - Whether or not the error should be printed during
training.
|
|
|
stopping_criteria
|
stopping_criteria (
self,
num_iterations,
training_error,
validation_error,
)
Define when to stop iterating.
|
|