def predict_stochastic(self, X, batch_size=128, verbose=0):
'''Generate output predictions for the input samples
batch by batch, using stochastic forward passes. If
dropout is used at training, during prediction network
units will be dropped at random as well. This procedure
can be used for MC dropout (see [ModelTest callbacks](callbacks.md)).
# Arguments
X: the input data, as a numpy array.
batch_size: integer.
verbose: verbosity mode, 0 or 1.
# Returns
A numpy array of predictions.
# References
- [Dropout: A simple way to prevent neural networks from overfitting](http://jmlr.org/papers/v15/srivastava14a.html)
- [Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning](http://arxiv.org/abs/1506.02142)
'''
X = models.standardize_X(X)
if self._predict_stochastic is None: # we only get self.model after init
self._predict_stochastic = K.function([self.model.X_test],
[self.model.y_train])
return self.model._predict_loop(self._predict_stochastic, X,
batch_size, verbose)[0]
def fit(self, X, batch_size=128, nb_epoch=100, verbose=1, callbacks=[],
validation_split=0., validation_data=None, shuffle=True, show_accuracy=False):
X = standardize_X(X)
val_f = None
val_ins = None
if validation_data or validation_split:
if show_accuracy:
val_f = self._test_with_acc
else:
val_f = self._test
if show_accuracy:
f = self._train_with_acc
out_labels = ['loss', 'acc']
else:
f = self._train
out_labels = ['loss']
ins = X# + [y, sample_weight]
metrics = ['loss', 'acc', 'val_loss', 'val_acc']
return self._fit(f, ins, out_labels=out_labels, batch_size=batch_size, nb_epoch=nb_epoch,
verbose=verbose, callbacks=callbacks,
val_f=val_f, val_ins=val_ins,
shuffle=shuffle, metrics=metrics)
def fit(self, X, batch_size=128, nb_epoch=100, verbose=1, callbacks=[],
validation_split=0., validation_data=None, shuffle=True, show_accuracy=False):
X = standardize_X(X)
# y = standardize_y(y)
val_f = None
val_ins = None
if validation_data or validation_split:
if show_accuracy:
val_f = self._test_with_acc
else:
val_f = self._test
# if validation_data:
# if len(validation_data) == 2:
# X_val, y_val = validation_data
# X_val = standardize_X(X_val)
# y_val = standardize_y(y_val)
# sample_weight_val = np.ones(y_val.shape[:-1] + (1,))
# elif len(validation_data) == 3:
# X_val, y_val, sample_weight_val = validation_data
# X_val = standardize_X(X_val)
# y_val = standardize_y(y_val)
# sample_weight_val = standardize_weights(y_val, sample_weight=sample_weight_val)
# else:
# raise Exception("Invalid format for validation data; provide a tuple (X_val, y_val) or (X_val, y_val, sample_weight). \
# X_val may be a numpy array or a list of numpy arrays depending on your model input.")
# val_ins = X_val + [y_val, sample_weight_val]
#
# elif 0 < validation_split < 1:
# split_at = int(len(X[0]) * (1 - validation_split))
# X, X_val = (slice_X(X, 0, split_at), slice_X(X, split_at))
# y, y_val = (slice_X(y, 0, split_at), slice_X(y, split_at))
# if sample_weight is not None:
# sample_weight, sample_weight_val = (slice_X(sample_weight, 0, split_at), slice_X(sample_weight, split_at))
# sample_weight_val = standardize_weights(y_val, sample_weight=sample_weight_val)
# else:
# sample_weight_val = np.ones(y_val.shape[:-1] + (1,))
# val_ins = X_val + [y_val, sample_weight_val]
if show_accuracy:
f = self._train_with_acc
out_labels = ['loss', 'acc']
else:
f = self._train
out_labels = ['loss']
ins = X# + [y, sample_weight]
metrics = ['loss', 'acc', 'val_loss', 'val_acc']
return self._fit(f, ins, out_labels=out_labels, batch_size=batch_size, nb_epoch=nb_epoch,
verbose=verbose, callbacks=callbacks,
val_f=val_f, val_ins=val_ins,
shuffle=shuffle, metrics=metrics)
def fit(self,
X,
batch_size=128,
nb_epoch=100,
verbose=1,
callbacks=[],
validation_split=0.,
validation_data=None,
shuffle=True,
show_accuracy=False):
X = standardize_X(X)
val_f = None
val_ins = None
if validation_data or validation_split:
if show_accuracy:
val_f = self._test_with_acc
else:
val_f = self._test
if show_accuracy:
f = self._train_with_acc
out_labels = ['loss', 'acc']
else:
f = self._train
out_labels = ['loss']
ins = X # + [y, sample_weight]
metrics = ['loss', 'acc', 'val_loss', 'val_acc']
return self._fit(f,
ins,
out_labels=out_labels,
batch_size=batch_size,
nb_epoch=nb_epoch,
verbose=verbose,
callbacks=callbacks,
val_f=val_f,
val_ins=val_ins,
shuffle=shuffle,
metrics=metrics)
def predict_stochastic(self, X, batch_size=128, verbose=0):
'''Generate output predictions for the input samples
batch by batch, using stochastic forward passes. If
dropout is used at training, during prediction network
units will be dropped at random as well. This procedure
can be used for MC dropout (see [ModelTest callbacks](callbacks.md)).
# Arguments
X: the input data, as a numpy array.
batch_size: integer.
verbose: verbosity mode, 0 or 1.
# Returns
A numpy array of predictions.
# References
- [Dropout: A simple way to prevent neural networks from overfitting](http://jmlr.org/papers/v15/srivastava14a.html)
- [Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning](http://arxiv.org/abs/1506.02142)
'''
X = models.standardize_X(X)
if self._predict_stochastic is None: # we only get self.model after init
self._predict_stochastic = K.function([self.model.X_test], [self.model.y_train])
return self.model._predict_loop(self._predict_stochastic, X, batch_size, verbose)[0]
def fit(self, X, batch_size=128, nb_epoch=100, verbose=1, callbacks=[],
validation_split=0., validation_data=None, shuffle=True, show_accuracy=False):
X = standardize_X(X)
# y = standardize_y(y)
# sample_weight = standardize_weights(y, class_weight=class_weight, sample_weight=sample_weight)
val_f = None
val_ins = None
if validation_data or validation_split:
if show_accuracy:
val_f = self._test_with_acc
else:
val_f = self._test
# if validation_data:
# try:
# X_val, y_val = validation_data
# except:
# raise Exception("Invalid format for validation data; provide a tuple (X_val, y_val). \
# X_val may be a numpy array or a list of numpy arrays depending on your model input.")
# X_val = standardize_X(X_val)
# y_val = standardize_y(y_val)
# val_ins = X_val + [y_val, np.ones(y_val.shape[:-1] + (1,))]
if show_accuracy:
f = self._train_with_acc
out_labels = ['loss', 'acc']
else:
f = self._train
out_labels = ['loss']
ins = X# + [y, sample_weight]
metrics = ['loss', 'acc', 'val_loss', 'val_acc']
return self._fit(f, ins, out_labels=out_labels, batch_size=batch_size, nb_epoch=nb_epoch,
verbose=verbose, callbacks=callbacks,
validation_split=validation_split, val_f=val_f, val_ins=val_ins,
shuffle=shuffle, metrics=metrics)
def train_on_batch(self, X):
X = standardize_X(X)
ins = X
return self._train(*ins)
def _predict(self, X, batch_size=128, verbose=0):
X = standardize_X(X)
return self._predict_loop(self._loss, X, batch_size, verbose)[0]
def _predict(self, X, batch_size=128, verbose=0):
X = standardize_X(X)
return self._predict_loop(self._loss, X, batch_size, verbose)[0]
