def inplace_logistic(X):
"""Compute the logistic function inplace.
Parameters
----------
X : {array-like, sparse matrix}, shape (n_samples, n_features)
The input data.
"""
logistic_sigmoid(X, out=X)
def predict(self, X):
"""Predict using the multi-layer perceptron model
Parameters
----------
X : {array-like, sparse matrix}, shape (n_samples, n_features)
Returns
-------
array, shape (n_samples)
Predicted target values per element in X.
"""
X = atleast2d_or_csr(X)
scores = self.decision_function(X)
if len(scores.shape) == 1 or self.multi_label is True:
scores = logistic_sigmoid(scores)
results = (scores > 0.5).astype(np.int)
if self.multi_label:
return self._lbin.inverse_transform(results)
else:
scores = _softmax(scores)
results = scores.argmax(axis=1)
return self.classes_[results]
def sigmoid(x):
'''激活函数，用于隐藏层输出
'''
#exp_x = np.exp(-x)
#sigmoid_x = 1 / (1 + exp_x)
sigmoid_x = logistic_sigmoid(x)
return sigmoid_x
def unigram_idf_mean_difference(entry):
entry['unigram_idf_mean_difference_feature'] = logistic_sigmoid(
abs(
geometric_mean_of_unigram_idfs(entry['question1_document']) -
geometric_mean_of_unigram_idfs(entry['question2_document'])
)
)
return entry
def test_logistic() -> None:
print('\ttest_logistic():')
X = np.concatenate((np.full((1, ), -np.inf), np.arange(-5, 5),
np.full((1, ), np.inf))).reshape(1, -1)
X_true = logistic_sigmoid(X)
ACTIVATIONS["logistic"](X)
np.testing.assert_array_equal(X, X_true)
X_true = np.negative(np.log(1 - X))
ACTIVATIONS_INVERSE["logistic"](X)
np.testing.assert_array_equal(X, X_true)
def logistic(X):
"""Compute the logistic function inplace.
Parameters
----------
X : {array-like, sparse matrix}, shape (n_samples, n_features)
The input data.
Returns
-------
X_new : {array-like, sparse matrix}, shape (n_samples, n_features)
The transformed data.
"""
return logistic_sigmoid(X, out=X)
def logistic(X):
"""Compute the logistic function inplace.
Parameters
----------
X : {array-like, sparse matrix}, shape (n_samples, n_features)
The input data.
Returns
-------
X_new : {array-like, sparse matrix}, shape (n_samples, n_features)
The transformed data.
"""
return logistic_sigmoid(X, out=X)
def predict_proba(self, X):
"""Probability estimates.
Parameters
----------
X : {array-like, sparse matrix}, shape (n_samples, n_features)
Returns
-------
array, shape (n_samples, n_outputs)
Returns the probability of the sample for each class in the model,
where classes are ordered as they are in `self.classes_`.
"""
scores = self.decision_function(X)
if len(scores.shape) == 1:
scores = logistic_sigmoid(scores)
return np.vstack([1 - scores, scores]).T
else:
return _softmax(scores)
def forward_propagate(self, word):
activations = []
input_to_layer = word
for layer in range(len(self.layers)-2):
input_to_layer = np.dot(input_to_layer,self.network.coefs_[layer])
input_to_layer += self.network.intercepts_[layer]
input_to_layer = np.maximum(input_to_layer, 0)
activations.append(input_to_layer)
input_to_layer = np.dot(input_to_layer, self.network.coefs_[-1])
input_to_layer += self.network.intercepts_[-1]
input_to_layer = logistic_sigmoid(input_to_layer)
activations.append(input_to_layer)
if self.verbose: print("Activations: {}".format(activations))
word = word.reshape(1,-1)
if self.verbose: print("Proba from net: {}".format(self.network.predict_proba(word)))
return activations
def _run(self, x): # pylint: disable=W0221
y = logistic_sigmoid(x)
return (y, )
def logistic(X):
return logistic_sigmoid(X, out=X)
lr=0.5,
train_split=None,
max_epochs=100)
lr_clf.initialize()
print("Weights:")
print(lr_clf.module_.linear_transform_layer.weight)
print("Bias:")
print(lr_clf.module_.linear_transform_layer.bias)
print("Random Data!")
# Generate random data
x_NF = np.random.randn(N, F)
true_w_F = np.arange(F) + 1
true_y_proba_N = logistic_sigmoid(np.dot(x_NF, true_w_F))
true_y_N = np.asarray(
true_y_proba_N >= np.random.rand(N),
dtype=np.float64)
clf = sklearn.linear_model.LogisticRegression(
C=0.5/lr_clf.l2_penalty_weights,
solver='lbfgs')
clf.fit(x_NF, true_y_N)
lr_clf.fit(x_NF, true_y_N)
print("EST BY SKORCH w:")
