def classifyTestSamples(trainingFeatures, trainingCategories, testFeatures):
clf = SGDClassifier()
clf.fit(trainingFeatures, trainingCategories)
predictedCategories = clf.predict(testFeatures)
return predictedCategories
class MachineLearning:
def __init__(self, Master_DF):
self.Data_Frame = Master_DF
def Encoder(self, df):
encoder = LabelEncoder()
print("Fitting")
encoder.fit(df)
return encoder.transform(df)
def Perceptorn_PreProcessing(self, x, y):
X = self.Encoder(self.Data_Frame[x].factorize()[0])
Y = self.Encoder(self.Data_Frame[y].factorize()[0])
X_train, X_test, Y_train, Y_test = train_test_split(X,
Y,
test_size=0.3,
random_state=0)
sc = StandardScaler()
sc.fit(X_train)
X_train_std = sc.transform(X_train)
X_test_std = sc.transform(X_test)
return X_train_std, Y_train, X_test_std, Y_test
def ppn_model(self, n_iter, eta0, random_state):
self.X_train_std_uri, self.Y_train_category, self.X_test_std_uri, self.Y_test_category = self.Perceptorn_PreProcessing(
'uri', 'category')
ppn = Perceptron(n_iter=n_iter, eta0=eta0, random_state=random_state)
ppn.fit(self.X_train_std_uri.reshape(len(self.X_train_std_uri), 1),
self.Y_train_category)
y_pred = ppn.predict(
self.X_test_std_uri.reshape(len(self.X_test_std_uri), 1))
return (self.Y_test_category != y_pred).sum(), accuracy_score(
self.Y_test_category, y_pred), y_pred
def lr_model(self, c, random_state):
lr = LogisticRegression(C=c, random_state=random_state)
lr.fit(self.X_train_std_uri.reshape(len(self.X_train_std_uri), 1),
self.Y_train_category)
y_pred = lr.predict(
self.X_test_std_uri.reshape(len(self.X_test_std_uri), 1))
return (self.Y_test_category != y_pred).sum(), accuracy_score(
self.Y_test_category, y_pred), y_pred
def ada_sd(self, n_iter, eta, random_state):
self.ada = SGDClassifier(n_iter=n_iter,
eta=eta,
random_state=random_state)
self.ada.fit(
self.X_train_std_uri.reshape(len(self.X_train_std_uri), 1),
self.Y_train_category)
self.y_pred = self.ada.predict(
self.X_test_std_uri.reshape(len(self.X_test_std_uri), 1))
return (self.Y_test_category != self.y_pred).sum(), accuracy_score(
self.Y_test_category, self.y_pred), self.y_pred
def withoutPipeline():
scaler = StandardScaler().fit(x_train) # for each x, (x - mean(all x))/std. dev. of x
# this step computes the mean and std. dev.
x_train = scaler.transform(x_train)
x_test = scaler.transform(x_test)
clfer = SGDClassifier()
clfer.fit(x_train, y_train) # this will try to separate the three classes based
# on the two features we gave it. Hence, we will get
# back three lines. I.e., three sets of coefficients
# and three intercepts
if(DEBUG):
#print clfer.coef_
#print clfer.intercept_
#print clfer.predict(scaler.transform([[4.7, 3.1]]))
#print clfer.decision_function(scaler.transform([[4.7, 3.1]])) # the algorithm evaluates distance from all three
# lines and picks the largest one (in this case [0])
pass
# validate results
y_predict_train = clfer.predict(x_train)
print "% Correct results on training set:"
print metrics.accuracy_score(y_train, y_predict_train)
y_predict_test = clfer.predict(x_test)
print "\n% Correct results on testing set:"
print metrics.accuracy_score(y_test, y_predict_test)
# Understanding the classification report:
# Precision: TP/(TP + FP) - ideal 1 - all instances reported as x were x. In other words,
# there were no instances reported as x that were NOT x
# Recall: TP/(TP + FN) - ideal 1 - all instances OF x were reported as x
# Although, accuracy does not appear in the report, it is important to know what it means:
# Accuracy: (TP + TN) / (TP + TN + FP + FN)
print "\nClassification Report:"
print metrics.classification_report(y_test, y_predict_test)
# Understanding the confusion matrix
# how many of class i were predicted as j
# ideal. an Identity matrix
print "Confusion Matrix:"
print metrics.confusion_matrix(y_test, y_predict_test)
class SGDClassifierImpl():
def __init__(self, loss='hinge', penalty='l2', alpha=0.0001, l1_ratio=0.15, fit_intercept=True, max_iter=None, tol=None, shuffle=True, verbose=0, epsilon=0.1, n_jobs=None, random_state=None, learning_rate='optimal', eta0=0.0, power_t=0.5, early_stopping=False, validation_fraction=0.1, n_iter_no_change=5, class_weight='balanced', warm_start=False, average=False):
self._hyperparams = {
'loss': loss,
'penalty': penalty,
'alpha': alpha,
'l1_ratio': l1_ratio,
'fit_intercept': fit_intercept,
'max_iter': max_iter,
'tol': tol,
'shuffle': shuffle,
'verbose': verbose,
'epsilon': epsilon,
'n_jobs': n_jobs,
'random_state': random_state,
'learning_rate': learning_rate,
'eta0': eta0,
'power_t': power_t,
'early_stopping': early_stopping,
'validation_fraction': validation_fraction,
'n_iter_no_change': n_iter_no_change,
'class_weight': class_weight,
'warm_start': warm_start,
'average': average}
self._wrapped_model = SKLModel(**self._hyperparams)
def fit(self, X, y=None):
if (y is not None):
self._wrapped_model.fit(X, y)
else:
self._wrapped_model.fit(X)
return self
def predict(self, X):
return self._wrapped_model.predict(X)
def predict_proba(self, X):
return self._wrapped_model.predict_proba(X)
def decision_function(self, X):
return self._wrapped_model.decision_function(X)
def partial_fit(self, X, y=None, classes = None):
if not hasattr(self, "_wrapped_model"):
self._wrapped_model = SKLModel(**self._hyperparams)
self._wrapped_model.partial_fit(X, y, classes = classes)
return self
def stochastic_descent(xtrain, ytrain, xtest):
clf = SGDClassifier(loss="hinge",
penalty="l2",
max_iter=10,
random_state=42,
alpha=1e-3,
tol=None)
print("SGD Fitting")
clf.fit(xtrain, ytrain)
# Saving the model with pickle
with open(base_dir + "Model", 'wb') as f:
pickle.dump(clf, f)
print("SGD Predicting")
ytest = clf.predict(xtest)
return ytest
n_iter=10)
trainloss = []
testloss = []
for i, chunk in enumerate(
pd.read_csv("cancer2.csv",
chunksize=chunksize,
header=None,
iterator=True)):
X = chunk.iloc[:, :-1]
y = chunk.iloc[:, -1]
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.33,
random_state=42)
estimator.partial_fit(X, y, classes=np.unique(y))
trainR2 = mean_squared_error(y_train, estimator.predict(X_train))
testR2 = mean_squared_error(y_test, estimator.predict(X_test))
trainloss.append(trainR2)
testloss.append(testR2)
print("trainloss:{:.4f},testloss:{:.4f} ".format(trainloss[-1],
testloss[-1]))
if i > 3:
break
# In[134]:
import matplotlib.pyplot as plt
plt.plot(trainloss)
plt.plot(testloss)
plt.legend(('train', 'test'))
plt.show()
data[column_names[10]],
test_size=0.25,
random_state=33)
#查验训练样本的数量和类别分布
print(y_train.value_counts())
#查验测试样本的数量和类别分布
print(y_test.value_counts())
#使用线性分类模型从事良/恶性乳腺癌肿瘤预测任务
#标准化数据，保证每个维度的特征数据方差为1，均值为0。使得预测结果不会被某些维度过大的特征值主导
ss = StandardScaler()
X_train = ss.fit_transform(X_train)
X_test = ss.transform(X_test)
#初始化 LogisticRegression 和 SGDClassifier
lr = LogisticRegression()
sgdc = SGDClassifier()
#调用 LogisticRegression 中的 fit 函数/模块用来训练模型参数
lr.fit(X_train, y_train)
#使用训练好的模型 lr 对 X_test进行预测，结果储存在变量 lr_y_predict 中
lr_y_predict = lr.predict(X_test)
#调用 SGDClassifier 中的 fit 函数/模块用来训练模型参数
sgdc.fit(X_train, y_train)
#使用训练好的模型 sgdc 对 X_test进行预测，结果储存在变量 sgdc_y_predict 中
sgdc_y_predict = sgdc.predict(X_test)
#使用线性分类模型从事良/恶性肿瘤预测任务的性能分析
#使用逻辑斯蒂回归模型自带的评分函数 score 获得模型在测试集上的准确性结果
print('LR分类器的精度：', lr.score(X_test, y_test))
#利用 classification_report 获得 LogisticRegression 其他三个指标的结果
print(classification_report(y_test, lr_y_predict, target_names=['良性', '恶性']))
class SGD(
IterativeComponentWithSampleWeight,
AutoSklearnClassificationAlgorithm,
):
def __init__(self, loss, penalty, alpha, fit_intercept, tol,
learning_rate, l1_ratio=0.15, epsilon=0.1,
eta0=0.01, power_t=0.5, average=False, random_state=None):
self.loss = loss
self.penalty = penalty
self.alpha = alpha
self.fit_intercept = fit_intercept
self.tol = tol
self.learning_rate = learning_rate
self.l1_ratio = l1_ratio
self.epsilon = epsilon
self.eta0 = eta0
self.power_t = power_t
self.random_state = random_state
self.average = average
self.estimator = None
def iterative_fit(self, X, y, n_iter=2, refit=False, sample_weight=None):
from sklearn.linear_model.stochastic_gradient import SGDClassifier
# Need to fit at least two iterations, otherwise early stopping will not
# work because we cannot determine whether the algorithm actually
# converged. The only way of finding this out is if the sgd spends less
# iterations than max_iter. If max_iter == 1, it has to spend at least
# one iteration and will always spend at least one iteration, so we
# cannot know about convergence.
if refit:
self.estimator = None
if self.estimator is None:
self.fully_fit_ = False
self.alpha = float(self.alpha)
self.l1_ratio = float(self.l1_ratio) if self.l1_ratio is not None \
else 0.15
self.epsilon = float(self.epsilon) if self.epsilon is not None \
else 0.1
self.eta0 = float(self.eta0)
self.power_t = float(self.power_t) if self.power_t is not None \
else 0.5
self.average = check_for_bool(self.average)
self.fit_intercept = check_for_bool(self.fit_intercept)
self.tol = float(self.tol)
self.estimator = SGDClassifier(loss=self.loss,
penalty=self.penalty,
alpha=self.alpha,
fit_intercept=self.fit_intercept,
max_iter=n_iter,
tol=self.tol,
learning_rate=self.learning_rate,
l1_ratio=self.l1_ratio,
epsilon=self.epsilon,
eta0=self.eta0,
power_t=self.power_t,
shuffle=True,
average=self.average,
random_state=self.random_state,
warm_start=True)
self.estimator.fit(X, y, sample_weight=sample_weight)
else:
self.estimator.max_iter += n_iter
self.estimator.max_iter = min(self.estimator.max_iter, 512)
self.estimator._validate_params()
self.estimator._partial_fit(
X, y,
alpha=self.estimator.alpha,
C=1.0,
loss=self.estimator.loss,
learning_rate=self.estimator.learning_rate,
max_iter=n_iter,
sample_weight=sample_weight,
classes=None,
coef_init=None,
intercept_init=None
)
if self.estimator._max_iter >= 512 or n_iter > self.estimator.n_iter_:
self.fully_fit_ = True
return self
def configuration_fully_fitted(self):
if self.estimator is None:
return False
elif not hasattr(self, 'fully_fit_'):
return False
else:
return self.fully_fit_
def predict(self, X):
if self.estimator is None:
raise NotImplementedError()
return self.estimator.predict(X)
def predict_proba(self, X):
if self.estimator is None:
raise NotImplementedError()
if self.loss in ["log", "modified_huber"]:
return self.estimator.predict_proba(X)
else:
df = self.estimator.decision_function(X)
return softmax(df)
@staticmethod
def get_properties(dataset_properties=None):
return {'shortname': 'SGD Classifier',
'name': 'Stochastic Gradient Descent Classifier',
'handles_regression': False,
'handles_classification': True,
'handles_multiclass': True,
'handles_multilabel': False,
'is_deterministic': True,
'input': (DENSE, SPARSE, UNSIGNED_DATA),
'output': (PREDICTIONS,)}
@staticmethod
def get_hyperparameter_search_space(dataset_properties=None):
cs = ConfigurationSpace()
loss = CategoricalHyperparameter("loss",
["hinge", "log", "modified_huber", "squared_hinge", "perceptron"],
default_value="log")
penalty = CategoricalHyperparameter(
"penalty", ["l1", "l2", "elasticnet"], default_value="l2")
alpha = UniformFloatHyperparameter(
"alpha", 1e-7, 1e-1, log=True, default_value=0.0001)
l1_ratio = UniformFloatHyperparameter(
"l1_ratio", 1e-9, 1, log=True, default_value=0.15)
fit_intercept = UnParametrizedHyperparameter("fit_intercept", "True")
tol = UniformFloatHyperparameter("tol", 1e-5, 1e-1, log=True,
default_value=1e-4)
epsilon = UniformFloatHyperparameter(
"epsilon", 1e-5, 1e-1, default_value=1e-4, log=True)
learning_rate = CategoricalHyperparameter(
"learning_rate", ["optimal", "invscaling", "constant"],
default_value="invscaling")
eta0 = UniformFloatHyperparameter(
"eta0", 1e-7, 1e-1, default_value=0.01, log=True)
power_t = UniformFloatHyperparameter("power_t", 1e-5, 1,
default_value=0.5)
average = CategoricalHyperparameter(
"average", ["False", "True"], default_value="False")
cs.add_hyperparameters([loss, penalty, alpha, l1_ratio, fit_intercept,
tol, epsilon, learning_rate, eta0, power_t,
average])
# TODO add passive/aggressive here, although not properly documented?
elasticnet = EqualsCondition(l1_ratio, penalty, "elasticnet")
epsilon_condition = EqualsCondition(epsilon, loss, "modified_huber")
power_t_condition = EqualsCondition(power_t, learning_rate,
"invscaling")
# eta0 is only relevant if learning_rate!='optimal' according to code
# https://github.com/scikit-learn/scikit-learn/blob/0.19.X/sklearn/
# linear_model/sgd_fast.pyx#L603
eta0_in_inv_con = InCondition(eta0, learning_rate, ["invscaling",
"constant"])
cs.add_conditions([elasticnet, epsilon_condition, power_t_condition,
eta0_in_inv_con])
return cs
https://blog.csdn.net/quiet_girl/article/details/72517053
"""
ss = StandardScaler()
X_train = ss.fit_transform(X_train) # 先拟合数据，然后转化它将其转化为标准形式
X_test = ss.transform(
X_test) # Perform standardization by centering and scaling（通过找中心和缩放等实现标准化）
# 初始化, Stochastic Gradient Descent Classifier & Logistic Regression
lr = LogisticRegression()
sgdc = SGDClassifier()
lr.fit(X_train, y_train) # LR分类器 训练模型
lr_y_predict = lr.predict(X_test) # 对X_test进行预测
sgdc.fit(X_train, y_train) # 随机梯度下降分类器
sgdc_y_predict = sgdc.predict(X_test) # 对X_test进行预测
# 性能分析（Performance）
print('Accuracy of LR Classfier:', lr.score(X_test, y_test))
print(
classification_report(y_test,
lr_y_predict,
target_names=['Benign', 'Malignant']))
print('Accuracy of SGD Classfier:', sgdc.score(X_test, y_test))
print(
classification_report(y_test,
sgdc_y_predict,
target_names=['Benign', 'Malignant']))
# sklearn分类：https://blog.csdn.net/u012526003/article/details/79054012
ss = StandardScaler()
x_train = ss.fit_transform(x_train)
x_test = ss.transform(x_test)
#初始化
lr = LogisticRegression()
sgdc = SGDClassifier()
#调用LogisticRegression中的fit函数/模块用来训练模型参数
lr.fit(x_train, y_train)
#使用训练好的模型Lr对x_test进行预测，结果存储在lr_y_predict中
lr_y_predict = lr.predict(x_test)
#调用SGDClassifier中的fit函数/模块来训练模型参数
sgdc.fit(x_train, y_train)
#使用训练好的模型sgdc对X_test进行预测，结果存储在变量sgdc_y_predict中
sgdc_y_predict = sgdc.predict(x_test)
from sklearn.metrics import classification_report
#使用评分函数score获得模型在测试集合上的准确性结果
print('Accuracy of LR Classifier:', lr.score(x_test, y_test))
#利用classification_report模块获得LogisticRegression其他三个指标的结果
print(
classification_report(y_test,
lr_y_predict,
target_names=['Benign', 'Malignant']))
print("\n")
print('Accuarcy of SGD Classifier:', sgdc.score(x_test, y_test))
print(
classification_report(y_test,
sgdc_y_predict,
class SGD(AutoSklearnClassificationAlgorithm):
def __init__(self,
loss,
penalty,
alpha,
fit_intercept,
n_iter,
learning_rate,
l1_ratio=0.15,
epsilon=0.1,
eta0=0.01,
power_t=0.5,
average=False,
random_state=None):
self.loss = loss
self.penalty = penalty
self.alpha = alpha
self.fit_intercept = fit_intercept
self.n_iter = n_iter
self.learning_rate = learning_rate
self.l1_ratio = l1_ratio
self.epsilon = epsilon
self.eta0 = eta0
self.power_t = power_t
self.random_state = random_state
self.average = average
self.estimator = None
def fit(self, X, y, sample_weight=None):
self.iterative_fit(X,
y,
n_iter=1,
sample_weight=sample_weight,
refit=True)
while not self.configuration_fully_fitted():
self.iterative_fit(X, y, n_iter=1, sample_weight=sample_weight)
return self
def iterative_fit(self, X, y, n_iter=1, refit=False, sample_weight=None):
from sklearn.linear_model.stochastic_gradient import SGDClassifier
if refit:
self.estimator = None
if self.estimator is None:
self.alpha = float(self.alpha)
self.fit_intercept = self.fit_intercept == 'True'
self.n_iter = int(self.n_iter)
self.l1_ratio = float(
self.l1_ratio) if self.l1_ratio is not None else 0.15
self.epsilon = float(
self.epsilon) if self.epsilon is not None else 0.1
self.eta0 = float(self.eta0)
self.power_t = float(
self.power_t) if self.power_t is not None else 0.25
self.average = self.average == 'True'
self.estimator = SGDClassifier(loss=self.loss,
penalty=self.penalty,
alpha=self.alpha,
fit_intercept=self.fit_intercept,
n_iter=n_iter,
learning_rate=self.learning_rate,
l1_ratio=self.l1_ratio,
epsilon=self.epsilon,
eta0=self.eta0,
power_t=self.power_t,
shuffle=True,
average=self.average,
random_state=self.random_state)
else:
self.estimator.n_iter += n_iter
self.estimator.partial_fit(X,
y,
classes=np.unique(y),
sample_weight=sample_weight)
if self.estimator.n_iter >= self.n_iter:
self.fully_fit_ = True
return self
def configuration_fully_fitted(self):
if self.estimator is None:
return False
elif not hasattr(self, 'fully_fit_'):
return False
else:
return self.fully_fit_
def predict(self, X):
if self.estimator is None:
raise NotImplementedError()
return self.estimator.predict(X)
def predict_proba(self, X):
if self.estimator is None:
raise NotImplementedError()
if self.loss in ["log", "modified_huber"]:
return self.estimator.predict_proba(X)
else:
df = self.estimator.decision_function(X)
return softmax(df)
@staticmethod
def get_properties(dataset_properties=None):
return {
'shortname': 'SGD Classifier',
'name': 'Stochastic Gradient Descent Classifier',
'handles_regression': False,
'handles_classification': True,
'handles_multiclass': True,
'handles_multilabel': False,
'is_deterministic': True,
'input': (DENSE, SPARSE, UNSIGNED_DATA),
'output': (PREDICTIONS, )
}
@staticmethod
def get_hyperparameter_search_space(dataset_properties=None):
cs = ConfigurationSpace()
loss = CategoricalHyperparameter(
"loss",
["hinge", "log", "modified_huber", "squared_hinge", "perceptron"],
default="log")
penalty = CategoricalHyperparameter("penalty",
["l1", "l2", "elasticnet"],
default="l2")
alpha = UniformFloatHyperparameter("alpha",
10e-7,
1e-1,
log=True,
default=0.0001)
l1_ratio = UniformFloatHyperparameter("l1_ratio",
1e-9,
1,
log=True,
default=0.15)
fit_intercept = UnParametrizedHyperparameter("fit_intercept", "True")
n_iter = UniformIntegerHyperparameter("n_iter",
5,
1000,
log=True,
default=20)
epsilon = UniformFloatHyperparameter("epsilon",
1e-5,
1e-1,
default=1e-4,
log=True)
learning_rate = CategoricalHyperparameter(
"learning_rate", ["optimal", "invscaling", "constant"],
default="optimal")
eta0 = UniformFloatHyperparameter("eta0", 10**-7, 0.1, default=0.01)
power_t = UniformFloatHyperparameter("power_t", 1e-5, 1, default=0.25)
average = CategoricalHyperparameter("average", ["False", "True"],
default="False")
cs.add_hyperparameters([
loss, penalty, alpha, l1_ratio, fit_intercept, n_iter, epsilon,
learning_rate, eta0, power_t, average
])
# TODO add passive/aggressive here, although not properly documented?
elasticnet = EqualsCondition(l1_ratio, penalty, "elasticnet")
epsilon_condition = EqualsCondition(epsilon, loss, "modified_huber")
# eta0 seems to be always active according to the source code; when
# learning_rate is set to optimial, eta0 is the starting value:
# https://github.com/scikit-learn/scikit-learn/blob/0.15.X/sklearn/linear_model/sgd_fast.pyx
#eta0_and_inv = EqualsCondition(eta0, learning_rate, "invscaling")
#eta0_and_constant = EqualsCondition(eta0, learning_rate, "constant")
#eta0_condition = OrConjunction(eta0_and_inv, eta0_and_constant)
power_t_condition = EqualsCondition(power_t, learning_rate,
"invscaling")
cs.add_conditions([elasticnet, epsilon_condition, power_t_condition])
return cs
[190,90,47],[175,64,39],[177,70,40],[159,55,37], [171,75,42],[181,85,43]] #Corresponding gender tags Y = ['male','female','female','female','male','male', 'male','female','male','female','male'] # Decision Tree classifier- takes in the input data to predict whether male or female classifier = tree.DecisionTreeClassifier() classifier = classifier.fit(X,Y) #Support Vector Machine Classifier classifier1 = svm.SVC() classifier1 = classifier1.fit(X,Y) #Stochastic Gradient Descent clf = SGDClassifier() clf = clf.fit(X,Y) #Prediction step #prediction for decision Trees prediction = classifier.predict([[172,75,35]]) print(prediction) #Prediction for Support Vector Machines prediction1 =classifier1.predict([[177,70,43]]) print(prediction1) #Prediction for Stochastic Gradient Descent pred = clf.predict([[172,75,35]]) print(pred)
X_batch_kernel_approx, y_batch_onehot = encode(X_batch, y_batch,
one_hot_encoder,
column_transformer,
rbf_sampler)
# make one pass of stochastic gradient descent over the batch.
sgd_classifier.partial_fit(X_batch_kernel_approx, y_batch, classes=[0, 1])
# print train/test accuracy metrics every 5 batch
if (batch_no % 5) == 0:
message = "batch {:>4} ".format(batch_no)
for origin, X, y_true_onehot in zip(
('train', 'val'), (X_batch_kernel_approx, X_test_kernel_approx),
(y_batch_onehot, y_true_test_onehot)):
y_pred = sgd_classifier.predict(X)
# preprocess correctly the labels and prediction to match
# average_precision_score expectations
y_pred_onehot = one_hot_encoder.transform(y_pred.reshape(-1, 1))
score = average_precision_score(y_true_onehot, y_pred_onehot)
message += "{} precision: {:.4f} ".format(origin, score)
if origin == 'val':
test_scores_rbf.append(score)
train_times_rbf.append(time.perf_counter() - t0)
online_train_set_sizes.append((batch_no + 1) * batchsize)
print(message)
###############################################################################
def run(keyn, nPart):
all_classes = np.array([0, 1])
allKeys = [l.split()[0] for l in open('keywordsAll.txt').readlines()]
keyFreqs = [
float(l.split()[1]) / 4205907
for l in open('keywordsAll.txt').readlines()
]
key = allKeys[keyn]
freq = keyFreqs[keyn]
opt = 'body+title+code'
bv = 'True'
nneg = 'True'
nv = 'None'
#testopt = 'c'
#testopt = 'w'
#testopt = 'l2'
testopt = 'l1'
if testopt == 'c':
cls = SGDClassifier(loss='hinge',
learning_rate="constant",
alpha=1e-6,
eta0=1e-2,
penalty='l2')
elif testopt == 'w':
cls = SGDClassifier(class_weight={1: 1.0 / freq / 8.0, 0: 1})
elif testopt == 'l2':
cls = SGDClassifier(loss='log', alpha=1e-5, penalty='l2')
elif testopt == 'l1':
cls = SGDClassifier(loss='log', alpha=1e-5, penalty='l1')
outputName = 'key_' + str(
keyn) + '_SGDtune_' + opt + '_partialfit_' + testopt + '.txt'
pklName = 'SGD_key_' + str(keyn) + '_' + testopt + '.pkl'
n0, ntrain = resumeJob(outputName, pklName)
body_test, y_test = getTestSet(10, key, opt, testSize=0.2, seed=123)
tot_pos = sum(y_test)
vectorizer = HashingVectorizer(decode_error='ignore',
n_features=2**20,
token_pattern=r"\b\w[\w#+.-]*(?<!\.$)",
binary=str2bool(bv),
norm=normOpt(nv),
non_negative=str2bool(nneg))
X_test = vectorizer.transform(body_test)
#print 'test case:', len(y_test), 'positive', tot_pos, 'key:', key, 'X norm:', X_test.sum(), 'binary:', bv, 'norm:', nv, 'nneg:', nneg
if n0 >= 2:
cls = joblib.load(pklName)
for n in xrange(n0, 10):
outfile = open(outputName, 'a')
data = json.load(gzip.open('Train.rdup.' + str(n) + '.json.gz'))
minibatch_size = len(data) / nPart + 1
for i in xrange(nPart):
n1 = i * minibatch_size
n2 = (i + 1) * minibatch_size
if i == nPart - 1:
n2 = len(data)
ntrain += (n2 - n1)
body_train, y_train = getMiniBatch(data, n1, n2, key, opt)
X_train = vectorizer.transform(body_train)
shuffledRange = range(n2 - n1)
for n_iter in xrange(5):
X_train, y_train = shuffle(X_train, y_train)
cls.partial_fit(X_train, y_train, classes=all_classes)
y_pred = cls.predict(X_test)
f1 = metrics.f1_score(y_test, y_pred)
p = metrics.precision_score(y_test, y_pred)
r = metrics.recall_score(y_test, y_pred)
accu = cls.score(X_train, y_train)
y_pred = cls.predict(X_train)
f1t = metrics.f1_score(y_train, y_pred)
outfile.write(
"%3d %8d %.4f %.3f %.3f %.3f %.3f %5d %5d\n" %
(n, ntrain, accu, f1t, f1, p, r, sum(y_pred), tot_pos))
_ = joblib.dump(cls, pklName, compress=9)
outfile.close()
class SGD(ParamSklearnClassificationAlgorithm):
def __init__(self, loss, penalty, alpha, fit_intercept, n_iter,
learning_rate, class_weight=None, l1_ratio=0.15, epsilon=0.1,
eta0=0.01, power_t=0.5, average=False, random_state=None):
self.loss = loss
self.penalty = penalty
self.alpha = alpha
self.fit_intercept = fit_intercept
self.n_iter = n_iter
self.learning_rate = learning_rate
self.class_weight = class_weight
self.l1_ratio = l1_ratio
self.epsilon = epsilon
self.eta0 = eta0
self.power_t = power_t
self.random_state = random_state
self.average = average
self.estimator = None
def fit(self, X, y):
while not self.configuration_fully_fitted():
self.iterative_fit(X, y, n_iter=1)
return self
def iterative_fit(self, X, y, n_iter=1, refit=False):
if refit:
self.estimator = None
if self.estimator is None:
self.alpha = float(self.alpha)
self.fit_intercept = self.fit_intercept == 'True'
self.n_iter = int(self.n_iter)
if self.class_weight == "None":
self.class_weight = None
self.l1_ratio = float(self.l1_ratio) if self.l1_ratio is not None else 0.15
self.epsilon = float(self.epsilon) if self.epsilon is not None else 0.1
self.eta0 = float(self.eta0)
self.power_t = float(self.power_t) if self.power_t is not None else 0.25
self.average = self.average == 'True'
self.estimator = SGDClassifier(loss=self.loss,
penalty=self.penalty,
alpha=self.alpha,
fit_intercept=self.fit_intercept,
n_iter=self.n_iter,
learning_rate=self.learning_rate,
class_weight=self.class_weight,
l1_ratio=self.l1_ratio,
epsilon=self.epsilon,
eta0=self.eta0,
power_t=self.power_t,
shuffle=True,
average=self.average,
random_state=self.random_state)
self.estimator.n_iter += n_iter
self.estimator.fit(X, y)
return self
def configuration_fully_fitted(self):
if self.estimator is None:
return False
return not self.estimator.n_iter < self.n_iter
def predict(self, X):
if self.estimator is None:
raise NotImplementedError()
return self.estimator.predict(X)
def predict_proba(self, X):
if self.estimator is None:
raise NotImplementedError()
if self.loss in ["log", "modified_huber"]:
return self.estimator.predict_proba(X)
else:
df = self.estimator.decision_function(X)
return softmax(df)
@staticmethod
def get_properties(dataset_properties=None):
return {'shortname': 'SGD Classifier',
'name': 'Stochastic Gradient Descent Classifier',
'handles_missing_values': False,
'handles_nominal_values': False,
'handles_numerical_features': True,
'prefers_data_scaled': True,
'prefers_data_normalized': True,
'handles_regression': False,
'handles_classification': True,
'handles_multiclass': True,
'handles_multilabel': False,
'is_deterministic': True,
'handles_sparse': True,
'input': (DENSE, SPARSE, UNSIGNED_DATA),
'output': (PREDICTIONS,),
# TODO find out what is best used here!
'preferred_dtype' : None}
@staticmethod
def get_hyperparameter_search_space(dataset_properties=None):
cs = ConfigurationSpace()
loss = cs.add_hyperparameter(CategoricalHyperparameter("loss",
["hinge", "log", "modified_huber", "squared_hinge", "perceptron"],
default="hinge"))
penalty = cs.add_hyperparameter(CategoricalHyperparameter(
"penalty", ["l1", "l2", "elasticnet"], default="l2"))
alpha = cs.add_hyperparameter(UniformFloatHyperparameter(
"alpha", 10e-7, 1e-1, log=True, default=0.0001))
l1_ratio = cs.add_hyperparameter(UniformFloatHyperparameter(
"l1_ratio", 0, 1, default=0.15))
fit_intercept = cs.add_hyperparameter(UnParametrizedHyperparameter(
"fit_intercept", "True"))
n_iter = cs.add_hyperparameter(UniformIntegerHyperparameter(
"n_iter", 5, 1000, default=20))
epsilon = cs.add_hyperparameter(UniformFloatHyperparameter(
"epsilon", 1e-5, 1e-1, default=1e-4, log=True))
learning_rate = cs.add_hyperparameter(CategoricalHyperparameter(
"learning_rate", ["optimal", "invscaling", "constant"],
default="optimal"))
eta0 = cs.add_hyperparameter(UniformFloatHyperparameter(
"eta0", 10**-7, 0.1, default=0.01))
power_t = cs.add_hyperparameter(UniformFloatHyperparameter(
"power_t", 1e-5, 1, default=0.25))
average = cs.add_hyperparameter(CategoricalHyperparameter(
"average", ["False", "True"], default="False"))
# TODO add passive/aggressive here, although not properly documented?
elasticnet = EqualsCondition(l1_ratio, penalty, "elasticnet")
epsilon_condition = EqualsCondition(epsilon, loss, "modified_huber")
# eta0 seems to be always active according to the source code; when
# learning_rate is set to optimial, eta0 is the starting value:
# https://github.com/scikit-learn/scikit-learn/blob/0.15.X/sklearn/linear_model/sgd_fast.pyx
#eta0_and_inv = EqualsCondition(eta0, learning_rate, "invscaling")
#eta0_and_constant = EqualsCondition(eta0, learning_rate, "constant")
#eta0_condition = OrConjunction(eta0_and_inv, eta0_and_constant)
power_t_condition = EqualsCondition(power_t, learning_rate, "invscaling")
cs.add_condition(elasticnet)
cs.add_condition(epsilon_condition)
cs.add_condition(power_t_condition)
return cs
def __str__(self):
return "ParamSklearn StochasticGradientClassifier"
class SGD(AutoSklearnClassificationAlgorithm):
def __init__(self,
loss,
penalty,
alpha,
fit_intercept,
n_iter,
learning_rate,
class_weight,
l1_ratio=0.15,
epsilon=0.1,
eta0=0.01,
power_t=0.5,
random_state=None):
self.loss = loss
self.penalty = penalty
self.alpha = alpha
self.fit_intercept = fit_intercept
self.n_iter = n_iter
self.learning_rate = learning_rate
self.class_weight = class_weight
self.l1_ratio = l1_ratio
self.epsilon = epsilon
self.eta0 = eta0
self.power_t = power_t
self.random_state = random_state
self.estimator = None
def fit(self, X, Y):
# TODO: maybe scale training data that its norm becomes 1?
# http://scikit-learn.org/stable/modules/sgd.html#id1
self.alpha = float(self.alpha)
self.fit_intercept = bool(self.fit_intercept)
self.n_iter = int(self.n_iter)
if self.class_weight == "None":
self.class_weight = None
self.l1_ratio = float(self.l1_ratio)
self.epsilon = float(self.epsilon)
self.eta0 = float(self.eta0)
self.power_t = float(self.power_t)
self.estimator = SGDClassifier(loss=self.loss,
penalty=self.penalty,
alpha=self.alpha,
fit_intercept=self.fit_intercept,
n_iter=self.n_iter,
learning_rate=self.learning_rate,
class_weight=self.class_weight,
l1_ratio=self.l1_ratio,
epsilon=self.epsilon,
eta0=self.eta0,
power_t=self.power_t,
shuffle=True,
random_state=self.random_state)
self.estimator.fit(X, Y)
return self
def predict(self, X):
if self.estimator is None:
raise NotImplementedError()
return self.estimator.predict(X)
def predict_proba(self, X):
if self.estimator is None:
raise NotImplementedError()
if self.loss in ["log", "modified_huber"]:
return self.estimator.predict_proba(X)
else:
df = self.estimator.decision_function(X)
return softmax(df)
@staticmethod
def get_properties():
return {
'shortname': 'SGD Classifier',
'name': 'Stochastic Gradient Descent Classifier',
'handles_missing_values': False,
'handles_nominal_values': False,
'handles_numerical_features': True,
'prefers_data_scaled': True,
'prefers_data_normalized': True,
'handles_multiclass': True,
'handles_multilabel': False,
'is_deterministic': True,
'handles_sparse': True,
# TODO find out what is best used here!
'preferred_dtype': None
}
@staticmethod
def get_hyperparameter_search_space(dataset_properties=None):
loss = CategoricalHyperparameter(
"loss",
["hinge", "log", "modified_huber", "squared_hinge", "perceptron"],
default="hinge")
penalty = CategoricalHyperparameter("penalty",
["l1", "l2", "elasticnet"],
default="l2")
alpha = UniformFloatHyperparameter("alpha",
10**-7,
10**-1,
log=True,
default=0.0001)
l1_ratio = UniformFloatHyperparameter("l1_ratio", 0, 1, default=0.15)
fit_intercept = UnParametrizedHyperparameter("fit_intercept", "True")
n_iter = UniformIntegerHyperparameter("n_iter", 5, 1000, default=20)
epsilon = UniformFloatHyperparameter("epsilon",
1e-5,
1e-1,
default=1e-4,
log=True)
learning_rate = CategoricalHyperparameter(
"learning_rate", ["optimal", "invscaling", "constant"],
default="optimal")
eta0 = UniformFloatHyperparameter("eta0", 10**-7, 0.1, default=0.01)
power_t = UniformFloatHyperparameter("power_t", 1e-5, 1, default=0.5)
# This does not allow for other resampling methods!
class_weight = CategoricalHyperparameter("class_weight",
["None", "auto"],
default="None")
cs = ConfigurationSpace()
cs.add_hyperparameter(loss)
cs.add_hyperparameter(penalty)
cs.add_hyperparameter(alpha)
cs.add_hyperparameter(l1_ratio)
cs.add_hyperparameter(fit_intercept)
cs.add_hyperparameter(n_iter)
cs.add_hyperparameter(epsilon)
cs.add_hyperparameter(learning_rate)
cs.add_hyperparameter(eta0)
cs.add_hyperparameter(power_t)
cs.add_hyperparameter(class_weight)
# TODO add passive/aggressive here, although not properly documented?
elasticnet = EqualsCondition(l1_ratio, penalty, "elasticnet")
epsilon_condition = EqualsCondition(epsilon, loss, "modified_huber")
# eta0 seems to be always active according to the source code; when
# learning_rate is set to optimial, eta0 is the starting value:
# https://github.com/scikit-learn/scikit-learn/blob/0.15.X/sklearn/linear_model/sgd_fast.pyx
#eta0_and_inv = EqualsCondition(eta0, learning_rate, "invscaling")
#eta0_and_constant = EqualsCondition(eta0, learning_rate, "constant")
#eta0_condition = OrConjunction(eta0_and_inv, eta0_and_constant)
power_t_condition = EqualsCondition(power_t, learning_rate,
"invscaling")
cs.add_condition(elasticnet)
cs.add_condition(epsilon_condition)
cs.add_condition(power_t_condition)
return cs
def __str__(self):
return "AutoSklearn StochasticGradientClassifier"
Listx = [[188, 57, 30], [167, 32, 22], [193, 65, 29], [185, 53, 27],
[164, 45, 22], [157, 38, 24], [179, 52, 27], [175, 68, 26],
[167, 39, 24], [178, 62, 27], [158, 46, 26]]
#List of labels Y. Gender female or male
Listy = [
'male', 'female', 'male', 'male', 'female', 'female', 'male', 'male',
'female', 'male', 'female'
]
#Model to store the decision tree model: Clasifier
Clasifier_tree = tree.DecisionTreeClassifier()
Clasifier_Sgradient = SGDClassifier()
Clasifier_naive = GaussianNB()
#Training stage
Clasifier_tree = Clasifier_tree.fit(Listx, Listy)
Clasifier_Sgradient = Clasifier_Sgradient.fit(Listx, Listy)
Clasifier_naive = Clasifier_naive.fit(Listx, Listy)
#Test stage
Listz = [[150, 35, 21]]
Prediction_tree = Clasifier_tree.predict(Listz)
Prediction_Gradient = Clasifier_Sgradient.predict(Listz)
Prediction_naive = Clasifier_naive.predict(Listz)
print(Prediction_tree)
print(Prediction_Gradient)
print(Prediction_naive)
class SGD:
def __init__(self,
loss,
penalty,
alpha,
fit_intercept,
tol,
learning_rate,
l1_ratio=0.15,
epsilon=0.1,
eta0=0.01,
power_t=0.5,
average=False,
random_state=None):
self.loss = loss
self.penalty = penalty
self.alpha = alpha
self.fit_intercept = fit_intercept
self.tol = tol
self.learning_rate = learning_rate
self.l1_ratio = l1_ratio
self.epsilon = epsilon
self.eta0 = eta0
self.power_t = power_t
self.random_state = random_state
self.average = average
self.estimator = None
def fit(self, X, y, sample_weight=None):
self.iterative_fit(X,
y,
n_iter=2,
refit=True,
sample_weight=sample_weight)
iteration = 2
while not self.configuration_fully_fitted():
n_iter = int(2**iteration / 2)
self.iterative_fit(X,
y,
n_iter=n_iter,
sample_weight=sample_weight)
iteration += 1
return self
def iterative_fit(self, X, y, n_iter=2, refit=False, sample_weight=None):
from sklearn.linear_model.stochastic_gradient import SGDClassifier
# Need to fit at least two iterations, otherwise early stopping will not
# work because we cannot determine whether the algorithm actually
# converged. The only way of finding this out is if the sgd spends less
# iterations than max_iter. If max_iter == 1, it has to spend at least
# one iteration and will always spend at least one iteration, so we
# cannot know about convergence.
if refit:
self.estimator = None
if self.estimator is None:
self.fully_fit_ = False
self.alpha = float(self.alpha)
self.l1_ratio = float(self.l1_ratio) if self.l1_ratio is not None \
else 0.15
self.epsilon = float(self.epsilon) if self.epsilon is not None \
else 0.1
self.eta0 = float(self.eta0)
self.power_t = float(self.power_t) if self.power_t is not None \
else 0.5
self.average = check_for_bool(self.average)
self.fit_intercept = check_for_bool(self.fit_intercept)
self.tol = float(self.tol)
self.estimator = SGDClassifier(loss=self.loss,
penalty=self.penalty,
alpha=self.alpha,
fit_intercept=self.fit_intercept,
max_iter=n_iter,
tol=self.tol,
learning_rate=self.learning_rate,
l1_ratio=self.l1_ratio,
epsilon=self.epsilon,
eta0=self.eta0,
power_t=self.power_t,
shuffle=True,
average=self.average,
random_state=self.random_state,
warm_start=True)
self.estimator.fit(X, y, sample_weight=sample_weight)
else:
self.estimator.max_iter += n_iter
self.estimator.max_iter = min(self.estimator.max_iter, 512)
self.estimator._validate_params()
self.estimator._partial_fit(
X,
y,
alpha=self.estimator.alpha,
C=1.0,
loss=self.estimator.loss,
learning_rate=self.estimator.learning_rate,
max_iter=n_iter,
sample_weight=sample_weight,
classes=None,
coef_init=None,
intercept_init=None)
if self.estimator._max_iter >= 512 or n_iter > self.estimator.n_iter_:
self.fully_fit_ = True
return self
def configuration_fully_fitted(self):
if self.estimator is None:
return False
elif not hasattr(self, 'fully_fit_'):
return False
else:
return self.fully_fit_
def predict(self, X):
if self.estimator is None:
raise NotImplementedError()
return self.estimator.predict(X)
def predict_proba(self, X):
if self.estimator is None:
raise NotImplementedError()
if self.loss in ["log", "modified_huber"]:
return self.estimator.predict_proba(X)
else:
df = self.estimator.decision_function(X)
return softmax(df)
class SGD(AutoSklearnClassificationAlgorithm):
def __init__(self,
loss,
penalty,
alpha,
fit_intercept,
tol,
learning_rate,
l1_ratio=0.15,
epsilon=0.1,
eta0=0.01,
power_t=0.5,
average=False,
random_state=None):
self.loss = loss
self.penalty = penalty
self.alpha = alpha
self.fit_intercept = fit_intercept
self.tol = tol
self.learning_rate = learning_rate
self.l1_ratio = l1_ratio
self.epsilon = epsilon
self.eta0 = eta0
self.power_t = power_t
self.random_state = random_state
self.average = average
self.estimator = None
def fit(self, X, y, sample_weight=None):
n_iter = 2
self.iterative_fit(X,
y,
n_iter=n_iter,
sample_weight=sample_weight,
refit=True)
while not self.configuration_fully_fitted():
n_iter *= 2
self.iterative_fit(X,
y,
n_iter=n_iter,
sample_weight=sample_weight)
return self
def iterative_fit(self, X, y, n_iter=2, refit=False, sample_weight=None):
from sklearn.linear_model.stochastic_gradient import SGDClassifier
# Need to fit at least two iterations, otherwise early stopping will not
# work because we cannot determine whether the algorithm actually
# converged. The only way of finding this out is if the sgd spends less
# iterations than max_iter. If max_iter == 1, it has to spend at least
# one iteration and will always spend at least one iteration, so we
# cannot know about convergence.
if refit:
self.estimator = None
if self.estimator is None:
self.alpha = float(self.alpha)
self.fit_intercept = self.fit_intercept == 'True'
self.l1_ratio = float(
self.l1_ratio) if self.l1_ratio is not None else 0.15
self.epsilon = float(
self.epsilon) if self.epsilon is not None else 0.1
self.eta0 = float(self.eta0)
self.power_t = float(
self.power_t) if self.power_t is not None else 0.25
self.average = self.average == 'True'
self.tol = float(self.tol)
self.estimator = SGDClassifier(loss=self.loss,
penalty=self.penalty,
alpha=self.alpha,
fit_intercept=self.fit_intercept,
max_iter=n_iter,
tol=self.tol,
learning_rate=self.learning_rate,
l1_ratio=self.l1_ratio,
epsilon=self.epsilon,
eta0=self.eta0,
power_t=self.power_t,
shuffle=True,
average=self.average,
random_state=self.random_state,
warm_start=True)
self.estimator.fit(X, y, sample_weight=sample_weight)
else:
self.estimator.max_iter += n_iter
self.estimator._validate_params()
self.estimator._partial_fit(
X,
y,
alpha=self.estimator.alpha,
C=1.0,
loss=self.estimator.loss,
learning_rate=self.estimator.learning_rate,
max_iter=n_iter,
sample_weight=sample_weight,
classes=None,
coef_init=None,
intercept_init=None)
if self.estimator._max_iter >= 1000 or n_iter > self.estimator.n_iter_:
self.fully_fit_ = True
return self
def configuration_fully_fitted(self):
if self.estimator is None:
return False
elif not hasattr(self, 'fully_fit_'):
return False
else:
return self.fully_fit_
def predict(self, X):
if self.estimator is None:
raise NotImplementedError()
return self.estimator.predict(X)
def predict_proba(self, X):
if self.estimator is None:
raise NotImplementedError()
if self.loss in ["log", "modified_huber"]:
return self.estimator.predict_proba(X)
else:
df = self.estimator.decision_function(X)
return softmax(df)
@staticmethod
def get_properties(dataset_properties=None):
return {
'shortname': 'SGD Classifier',
'name': 'Stochastic Gradient Descent Classifier',
'handles_regression': False,
'handles_classification': True,
'handles_multiclass': True,
'handles_multilabel': False,
'is_deterministic': True,
'input': (DENSE, SPARSE, UNSIGNED_DATA),
'output': (PREDICTIONS, )
}
@staticmethod
def get_hyperparameter_search_space(dataset_properties=None):
cs = ConfigurationSpace()
loss = CategoricalHyperparameter(
"loss",
["hinge", "log", "modified_huber", "squared_hinge", "perceptron"],
default_value="log")
penalty = CategoricalHyperparameter("penalty",
["l1", "l2", "elasticnet"],
default_value="l2")
alpha = UniformFloatHyperparameter("alpha",
1e-7,
1e-1,
log=True,
default_value=0.0001)
l1_ratio = UniformFloatHyperparameter("l1_ratio",
1e-9,
1,
log=True,
default_value=0.15)
fit_intercept = UnParametrizedHyperparameter("fit_intercept", "True")
tol = UniformFloatHyperparameter("tol",
1e-5,
1e-1,
log=True,
default_value=1e-4)
epsilon = UniformFloatHyperparameter("epsilon",
1e-5,
1e-1,
default_value=1e-4,
log=True)
learning_rate = CategoricalHyperparameter(
"learning_rate", ["optimal", "invscaling", "constant"],
default_value="invscaling")
eta0 = UniformFloatHyperparameter("eta0",
1e-7,
1e-1,
default_value=0.01)
power_t = UniformFloatHyperparameter("power_t",
1e-5,
1,
default_value=0.25)
average = CategoricalHyperparameter("average", ["False", "True"],
default_value="False")
cs.add_hyperparameters([
loss, penalty, alpha, l1_ratio, fit_intercept, tol, epsilon,
learning_rate, eta0, power_t, average
])
# TODO add passive/aggressive here, although not properly documented?
elasticnet = EqualsCondition(l1_ratio, penalty, "elasticnet")
epsilon_condition = EqualsCondition(epsilon, loss, "modified_huber")
# eta0 seems to be always active according to the source code; when
# learning_rate is set to optimial, eta0 is the starting value:
# https://github.com/scikit-learn/scikit-learn/blob/0.15.X/sklearn/linear_model/sgd_fast.pyx
#eta0_and_inv = EqualsCondition(eta0, learning_rate, "invscaling")
#eta0_and_constant = EqualsCondition(eta0, learning_rate, "constant")
#eta0_condition = OrConjunction(eta0_and_inv, eta0_and_constant)
power_t_condition = EqualsCondition(power_t, learning_rate,
"invscaling")
cs.add_conditions([elasticnet, epsilon_condition, power_t_condition])
return cs
X_train_feature = vec.fit_transform(train_data['word_seg'])
X_test_feature = vec.transform(test_data['word_seg'])
# --------------情感值预测开始------------------------
y_train_sent = train_data['sentiment_value'].astype(int)
X_train_sent,X_test_sent,y_train_sent,y_test_sent=\
train_test_split(X_train_feature,y_train_sent,test_size=0.1,random_state=42)
# clf = LogisticRegression(C=4, dual=True)
# clf =svm.LinearSVC()
# clf =RandomForestClassifier()
clf = SGDClassifier(n_iter=80)
# tune_params(X_train_sent,y_train_sent)
clf.fit(X_train_sent, y_train_sent)
# 在训练集评估模型
pred_test_sent = clf.predict(X_test_sent)
# 精确度=真阳性/（真阳性+假阳性）
precision = precision_score(y_test_sent,
pred_test_sent,
pos_label=None,
average='weighted')
# 召回率=真阳性/（真阳性+假阴性）
recall = recall_score(y_test_sent,
pred_test_sent,
pos_label=None,
average='weighted')
# F1
f1 = f1_score(y_test_sent, pred_test_sent, pos_label=None, average='weighted')
# 精确率
accuracy = accuracy_score(y_test_sent, pred_test_sent)
print("precision:{:.4f}-recall:{:.4f}-f1:{:.4f}-accuracy:{:.4f}".format(
# sklearn implementation
from sklearn.linear_model import Perceptron
from sklearn.linear_model.stochastic_gradient import SGDClassifier
from sklearn.metrics import accuracy_score
# Fitting an sklearn Perceptron and SGDClassifier with perceptron loss function (these should be identical)
clf = Perceptron(random_state=None, eta0= 0.1, shuffle=False, penalty=None, class_weight=None, fit_intercept=False)
clf2 = SGDClassifier(loss="perceptron",eta0=0.1,learning_rate="constant",penalty=None,random_state=None,shuffle=False,fit_intercept=False,warm_start=False,average=False,n_iter=1000)
clf.fit(x_train, y_train)
clf2.fit(x_train, y_train)
y_predict = clf.predict(x_test)
y_preSGD = clf2.predict(x_test)
print "sklearn Perceptron accuracy:"
print accuracy_score(y_test, y_predict)
print "sklearn SGDClassifier accuracy:"
print accuracy_score(y_test, y_preSGD)
print "my perceptron accuracy:"
print accuracy_score(y_test, y_pred)
print "\n"
#print clf.coef_
def x22(x1, w):
w0 = clf.coef_[0][0]
w1 = clf.coef_[0][1]
class SGD(
IterativeComponentWithSampleWeight,
BaseClassificationModel,
):
def __init__(self, loss, penalty, alpha, fit_intercept, tol,
learning_rate, l1_ratio=0.15, epsilon=0.1,
eta0=0.01, power_t=0.5, average=False, random_state=None):
self.loss = loss
self.penalty = penalty
self.alpha = alpha
self.fit_intercept = fit_intercept
self.tol = tol
self.learning_rate = learning_rate
self.l1_ratio = l1_ratio
self.epsilon = epsilon
self.eta0 = eta0
self.power_t = power_t
self.random_state = random_state
self.average = average
self.estimator = None
self.time_limit = None
self.start_time = time.time()
def iterative_fit(self, X, y, n_iter=2, refit=False, sample_weight=None):
from sklearn.linear_model.stochastic_gradient import SGDClassifier
# Need to fit at least two iterations, otherwise early stopping will not
# work because we cannot determine whether the algorithm actually
# converged. The only way of finding this out is if the sgd spends less
# iterations than max_iter. If max_iter == 1, it has to spend at least
# one iteration and will always spend at least one iteration, so we
# cannot know about convergence.
if refit:
self.estimator = None
if self.estimator is None:
if isinstance(self.loss, tuple):
nested_loss = self.loss
self.loss = nested_loss[0]
if self.loss == 'modified_huber':
self.epsilon = nested_loss[1]['epsilon']
if isinstance(self.penalty, tuple):
nested_penalty = self.penalty
self.penalty = nested_penalty[0]
if self.penalty == "elasticnet":
self.l1_ratio = nested_penalty[1]['l1_ratio']
if isinstance(self.learning_rate, tuple):
nested_learning_rate = self.learning_rate
self.learning_rate = nested_learning_rate[0]
if self.learning_rate == 'invscaling':
self.eta0 = nested_learning_rate[1]['eta0']
self.power_t = nested_learning_rate[1]['power_t']
elif self.learning_rate == 'constant':
self.eta0 = nested_learning_rate[1]['eta0']
self.fully_fit_ = False
self.alpha = float(self.alpha)
self.l1_ratio = float(self.l1_ratio) if self.l1_ratio is not None \
else 0.15
self.epsilon = float(self.epsilon) if self.epsilon is not None \
else 0.1
self.eta0 = float(self.eta0) if self.eta0 is not None else 0.01
self.power_t = float(self.power_t) if self.power_t is not None \
else 0.5
self.average = check_for_bool(self.average)
self.fit_intercept = check_for_bool(self.fit_intercept)
self.tol = float(self.tol)
self.estimator = SGDClassifier(loss=self.loss,
penalty=self.penalty,
alpha=self.alpha,
fit_intercept=self.fit_intercept,
max_iter=n_iter,
tol=self.tol,
learning_rate=self.learning_rate,
l1_ratio=self.l1_ratio,
epsilon=self.epsilon,
eta0=self.eta0,
power_t=self.power_t,
shuffle=True,
average=self.average,
random_state=self.random_state,
warm_start=True)
self.estimator.fit(X, y, sample_weight=sample_weight)
else:
self.estimator.max_iter += n_iter
self.estimator.max_iter = min(self.estimator.max_iter, 512)
self.estimator._validate_params()
self.estimator._partial_fit(
X, y,
alpha=self.estimator.alpha,
C=1.0,
loss=self.estimator.loss,
learning_rate=self.estimator.learning_rate,
max_iter=n_iter,
sample_weight=sample_weight,
classes=None,
coef_init=None,
intercept_init=None
)
if self.estimator.max_iter >= 512 or n_iter > self.estimator.n_iter_:
self.fully_fit_ = True
return self
def configuration_fully_fitted(self):
if self.estimator is None:
return False
elif not hasattr(self, 'fully_fit_'):
return False
else:
return self.fully_fit_
def predict(self, X):
if self.estimator is None:
raise NotImplementedError()
return self.estimator.predict(X)
def predict_proba(self, X):
if self.estimator is None:
raise NotImplementedError()
if self.loss in ["log", "modified_huber"]:
return self.estimator.predict_proba(X)
else:
df = self.estimator.decision_function(X)
return softmax(df)
@staticmethod
def get_properties(dataset_properties=None):
return {'shortname': 'SGD Classifier',
'name': 'Stochastic Gradient Descent Classifier',
'handles_regression': False,
'handles_classification': True,
'handles_multiclass': True,
'handles_multilabel': False,
'is_deterministic': True,
'input': (DENSE, SPARSE, UNSIGNED_DATA),
'output': (PREDICTIONS,)}
@staticmethod
def get_hyperparameter_search_space(dataset_properties=None, optimizer='smac'):
if optimizer == 'smac':
cs = ConfigurationSpace()
loss = CategoricalHyperparameter("loss",
["hinge", "log", "modified_huber", "squared_hinge", "perceptron"],
default_value="log")
penalty = CategoricalHyperparameter(
"penalty", ["l1", "l2", "elasticnet"], default_value="l2")
alpha = UniformFloatHyperparameter(
"alpha", 1e-7, 1e-1, log=True, default_value=0.0001)
l1_ratio = UniformFloatHyperparameter(
"l1_ratio", 1e-9, 1, log=True, default_value=0.15)
fit_intercept = UnParametrizedHyperparameter("fit_intercept", "True")
tol = UniformFloatHyperparameter("tol", 1e-5, 1e-1, log=True,
default_value=1e-4)
epsilon = UniformFloatHyperparameter(
"epsilon", 1e-5, 1e-1, default_value=1e-4, log=True)
learning_rate = CategoricalHyperparameter(
"learning_rate", ["optimal", "invscaling", "constant"],
default_value="invscaling")
eta0 = UniformFloatHyperparameter(
"eta0", 1e-7, 1e-1, default_value=0.01, log=True)
power_t = UniformFloatHyperparameter("power_t", 1e-5, 1, log=True,
default_value=0.5)
average = CategoricalHyperparameter(
"average", ["False", "True"], default_value="False")
cs.add_hyperparameters([loss, penalty, alpha, l1_ratio, fit_intercept,
tol, epsilon, learning_rate, eta0, power_t,
average])
# TODO add passive/aggressive here, although not properly documented?
elasticnet = EqualsCondition(l1_ratio, penalty, "elasticnet")
epsilon_condition = EqualsCondition(epsilon, loss, "modified_huber")
power_t_condition = EqualsCondition(power_t, learning_rate,
"invscaling")
# eta0 is only relevant if learning_rate!='optimal' according to code
# https://github.com/scikit-learn/scikit-learn/blob/0.19.X/sklearn/
# linear_model/sgd_fast.pyx#L603
eta0_in_inv_con = InCondition(eta0, learning_rate, ["invscaling",
"constant"])
cs.add_conditions([elasticnet, epsilon_condition, power_t_condition,
eta0_in_inv_con])
return cs
elif optimizer == 'tpe':
eta0 = hp.loguniform('sgd_eta0', np.log(1e-7), np.log(1e-1))
space = {
'loss': hp.choice('sgd_loss', [
("modified_huber", {'epsilon': hp.loguniform('sgd_epsilon', np.log(1e-5), np.log(1e-1))}),
("hinge", {}),
("log", {}),
("squared_hinge", {}),
("perceptron", {})]),
'penalty': hp.choice('sgd_penalty',
[("elasticnet",
{'l1_ratio': hp.loguniform('sgd_l1_ratio', np.log(1e-9), np.log(1))}),
("l1", None),
("l2", None)]),
'alpha': hp.loguniform('sgd_alpha', np.log(1e-7), np.log(1e-1)),
'fit_intercept': hp.choice('sgd_fit_intercept', ["True"]),
'tol': hp.loguniform('sgd_tol', np.log(1e-5), np.log(1e-1)),
'learning_rate': hp.choice('sgd_learning_rate', [("optimal", {}),
("invscaling",
{'power_t': hp.loguniform('sgd_power_t', np.log(1e-5),
np.log(1)),
'eta0': eta0}),
("constant", {'eta0': eta0})]),
'average': hp.choice('sgd_average', ["True", "False"])}
init_trial = {'loss': ("log", {}),
'penalty': ("l2", {}),
'alpha': 1e-4,
'fit_intercept': "True",
'tol': 1e-4,
'learning_rate': ("invscaling", {'power_t': 0.5, 'eta0': 0.01}),
'average': "False"}
return space
def result():
if request.method == 'POST':
path = request.files.get('myFile')
# Reading the CSV file and converting it into a pandas data-frame
df = pd.read_csv(path, encoding="ISO-8859-1")
# Reading the name for the file for the model that will be saved
filename = request.form['filename']
# Reading the names of the feature and label as strings
str1 = request.form['feature']
str2 = request.form['label']
# Assigning the feature and label variables to the respective columns
if str1 in list(df) and str2 in list(df):
y = df[str2]
X = df[str1]
else:
return render_template('nameError.html')
'''
# Removing the punctuations and HTTP links in the feature text input
x = []
for subject in X:
result = re.sub(r"http\S+", "", subject)
replaced = re.sub(r'[^a-zA-Z0-9 ]+', '', result)
x.append(replaced)
X = pd.Series(x)
'''
X = X.str.lower()
# Optional use of Tokenization and Lemmatization using Natural Language Processing in SpaCy
"""
texts = []
for doc in X:
doc = nlp(doc, disable=['parser', 'ner'])
tokens = [tok.lemma_.lower().strip() for tok in doc if tok.lemma_ != '-PRON-']
tokens = [tok for tok in tokens if tok not in stopwords]
tokens = ' '.join(tokens)
texts.append(tokens)
X = pd.Series(texts)
"""
# Splitting the data-set into 2 parts : Training data and Test data
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.33,
shuffle=True)
tfidfvect = TfidfVectorizer(ngram_range=(1, 1))
X_train_tfidf = tfidfvect.fit_transform(X_train)
# Fitting all the classification models one by one and recording their accuracies and execution times
start = time()
clf1 = LinearSVC()
clf1.fit(X_train_tfidf, y_train)
pred_SVC = clf1.predict(tfidfvect.transform(X_test))
a1 = accuracy_score(y_test, pred_SVC)
end = time()
print("accuracy SVC: {} and time: {} s".format(a1, (end - start)))
start = time()
clf2 = LogisticRegression(n_jobs=-1,
multi_class='multinomial',
solver='newton-cg')
clf2.fit(X_train_tfidf, y_train)
pred_LR = clf2.predict(tfidfvect.transform(X_test))
a2 = accuracy_score(y_test, pred_LR)
end = time()
print("accuracy LR: {} and time: {}".format(a2, (end - start)))
start = time()
clf3 = RandomForestClassifier(n_jobs=-1)
clf3.fit(X_train_tfidf, y_train)
pred = clf3.predict(tfidfvect.transform(X_test))
a3 = accuracy_score(y_test, pred)
end = time()
print("accuracy RFC: {} and time: {}".format(a3, (end - start)))
start = time()
clf4 = MultinomialNB()
clf4.fit(X_train_tfidf, y_train)
pred = clf4.predict(tfidfvect.transform(X_test))
a4 = accuracy_score(y_test, pred)
end = time()
print("accuracy MNB: {} and time: {}".format(a4, (end - start)))
start = time()
clf11 = SGDClassifier(n_jobs=-1)
clf11.fit(X_train_tfidf, y_train)
pred = clf11.predict(tfidfvect.transform(X_test))
a11 = accuracy_score(y_test, pred)
end = time()
print("accuracy SGDC: {} and time: {}".format(a11, (end - start)))
start = time()
clf12 = SGDClassifier(n_jobs=-1)
clf12.fit(X_train_tfidf, y_train)
pred = clf12.predict(tfidfvect.transform(X_test))
a12 = accuracy_score(y_test, pred)
end = time()
print("accuracy XGBC: {} and time: {}".format(a12, (end - start)))
# Comparing the accuracies of all the models and then saving(dumping) the model with the highest accuracy using pickle for later use.
acu_list = [a1, a2, a3, a4, a11, a12]
max_list = max(acu_list)
if max_list == a1:
pickle.dump(clf1, open(filename + '_model', 'wb'))
elif max_list == a2:
pickle.dump(clf2, open(filename + '_model', 'wb'))
elif max_list == a3:
pickle.dump(clf3, open(filename + '_model', 'wb'))
elif max_list == a4:
pickle.dump(clf4, open(filename + '_model', 'wb'))
elif max_list == a11:
pickle.dump(clf11, open(filename + '_model', 'wb'))
elif max_list == a12:
pickle.dump(clf12, open(filename + '_model', 'wb'))
pickle.dump(tfidfvect, open(filename + '_tfidfVect', 'wb'))
return render_template("result.html",
ac1=a1,
ac2=a2,
ac3=a3,
ac4=a4,
ac11=a11,
ac12=a12)
def result():
if request.method == 'POST':
path = request.files.get('myFile')
df = pd.read_csv(path, encoding="ISO-8859-1")
filename = request.form['filename']
str1 = request.form['feature']
str2 = request.form['label']
if str1 in list(df) and str2 in list(df):
y = df[str2]
X = df[str1]
else:
return render_template('nameError.html')
x = []
for subject in X:
result = re.sub(r"http\S+", "", subject)
replaced = re.sub(r'[^a-zA-Z0-9 ]+', '', result)
x.append(replaced)
X = pd.Series(x)
X = X.str.lower()
"""
texts = []
for doc in X:
doc = nlp(doc, disable=['parser', 'ner'])
tokens = [tok.lemma_.lower().strip() for tok in doc if tok.lemma_ != '-PRON-']
tokens = [tok for tok in tokens if tok not in stopwords]
tokens = ' '.join(tokens)
texts.append(tokens)
X = pd.Series(texts)
"""
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.33)
tfidfvect = TfidfVectorizer(ngram_range=(1, 1))
X_train_tfidf = tfidfvect.fit_transform(X_train)
start = time()
clf1 = LinearSVC()
clf1.fit(X_train_tfidf, y_train)
pred_SVC = clf1.predict(tfidfvect.transform(X_test))
a1 = accuracy_score(y_test, pred_SVC)
end = time()
print("accuracy SVC: {} and time: {} s".format(a1, (end - start)))
start = time()
clf2 = LogisticRegression(n_jobs=-1,
multi_class='multinomial',
solver='newton-cg')
clf2.fit(X_train_tfidf, y_train)
pred_LR = clf2.predict(tfidfvect.transform(X_test))
a2 = accuracy_score(y_test, pred_LR)
end = time()
print("accuracy LR: {} and time: {}".format(a2, (end - start)))
start = time()
clf3 = RandomForestClassifier(n_jobs=-1)
clf3.fit(X_train_tfidf, y_train)
pred = clf3.predict(tfidfvect.transform(X_test))
a3 = accuracy_score(y_test, pred)
end = time()
print("accuracy RFC: {} and time: {}".format(a3, (end - start)))
start = time()
clf4 = MultinomialNB()
clf4.fit(X_train_tfidf, y_train)
pred = clf4.predict(tfidfvect.transform(X_test))
a4 = accuracy_score(y_test, pred)
end = time()
print("accuracy MNB: {} and time: {}".format(a4, (end - start)))
start = time()
clf5 = GaussianNB()
clf5.fit(X_train_tfidf.toarray(), y_train)
pred = clf5.predict(tfidfvect.transform(X_test).toarray())
a5 = accuracy_score(y_test, pred)
end = time()
print("accuracy GNB: {} and time: {}".format(a5, (end - start)))
start = time()
clf6 = LogisticRegressionCV(n_jobs=-1)
clf6.fit(X_train_tfidf, y_train)
pred_LR = clf6.predict(tfidfvect.transform(X_test))
a6 = accuracy_score(y_test, pred_LR)
end = time()
print("accuracy LRCV: {} and time: {}".format(a6, (end - start)))
start = time()
clf7 = AdaBoostClassifier()
clf7.fit(X_train_tfidf, y_train)
pred_LR = clf7.predict(tfidfvect.transform(X_test))
a7 = accuracy_score(y_test, pred_LR)
end = time()
print("accuracy ABC: {} and time: {}".format(a7, (end - start)))
start = time()
clf8 = BernoulliNB()
clf8.fit(X_train_tfidf.toarray(), y_train)
pred = clf8.predict(tfidfvect.transform(X_test).toarray())
a8 = accuracy_score(y_test, pred)
end = time()
print("accuracy BNB: {} and time: {}".format(a8, (end - start)))
start = time()
clf9 = Perceptron(n_jobs=-1)
clf9.fit(X_train_tfidf.toarray(), y_train)
pred = clf9.predict(tfidfvect.transform(X_test).toarray())
a9 = accuracy_score(y_test, pred)
end = time()
print("accuracy Per: {} and time: {}".format(a9, (end - start)))
start = time()
clf10 = RidgeClassifierCV()
clf10.fit(X_train_tfidf.toarray(), y_train)
pred = clf10.predict(tfidfvect.transform(X_test).toarray())
a10 = accuracy_score(y_test, pred)
end = time()
print("accuracy RidCV: {} and time: {}".format(a10, (end - start)))
start = time()
clf11 = SGDClassifier(n_jobs=-1)
clf11.fit(X_train_tfidf.toarray(), y_train)
pred = clf11.predict(tfidfvect.transform(X_test).toarray())
a11 = accuracy_score(y_test, pred)
end = time()
print("accuracy SGDC: {} and time: {}".format(a11, (end - start)))
start = time()
clf12 = SGDClassifier(n_jobs=-1)
clf12.fit(X_train_tfidf.toarray(), y_train)
pred = clf12.predict(tfidfvect.transform(X_test).toarray())
a12 = accuracy_score(y_test, pred)
end = time()
print("accuracy XGBC: {} and time: {}".format(a12, (end - start)))
acu_list = [a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12]
max_list = max(acu_list)
if max_list == a1:
pickle.dump(clf1, open(filename + '_model', 'wb'))
elif max_list == a2:
pickle.dump(clf2, open(filename + '_model', 'wb'))
elif max_list == a3:
pickle.dump(clf3, open(filename + '_model', 'wb'))
elif max_list == a4:
pickle.dump(clf4, open(filename + '_model', 'wb'))
elif max_list == a5:
pickle.dump(clf5, open(filename + '_model', 'wb'))
elif max_list == a6:
pickle.dump(clf6, open(filename + '_model', 'wb'))
elif max_list == a7:
pickle.dump(clf7, open(filename + '_model', 'wb'))
elif max_list == a8:
pickle.dump(clf8, open(filename + '_model', 'wb'))
elif max_list == a9:
pickle.dump(clf9, open(filename + '_model', 'wb'))
elif max_list == a10:
pickle.dump(clf10, open(filename + '_model', 'wb'))
elif max_list == a11:
pickle.dump(clf11, open(filename + '_model', 'wb'))
elif max_list == a12:
pickle.dump(clf12, open(filename + '_model', 'wb'))
pickle.dump(tfidfvect, open(filename + '_tfidfVect', 'wb'))
return render_template("result.html",
ac1=a1,
ac2=a2,
ac3=a3,
ac4=a4,
ac5=a5,
ac6=a6,
ac7=a7,
ac8=a8,
ac9=a9,
ac10=a10,
ac11=a11,
ac12=a12)
# sklearn implementation
from sklearn.linear_model import Perceptron
from sklearn.linear_model.stochastic_gradient import SGDClassifier
from sklearn.metrics import accuracy_score
# Fitting an sklearn Perceptron and SGDClassifier with perceptron loss function (these should be identical)
clf = Perceptron(random_state=None, eta0= 0.1, shuffle=False, penalty=None, class_weight=None, fit_intercept=False)
clf2 = SGDClassifier(loss="perceptron",eta0=0.1,learning_rate="constant",penalty=None,random_state=None,shuffle=False,fit_intercept=False,warm_start=False,average=False,n_iter=1000)
clf.fit(x_train, y_train)
clf2.fit(x_train, y_train)
y_predict = clf.predict(x_test)
y_preSGD = clf2.predict(x_test)
print "sklearn Perceptron accuracy:"
print accuracy_score(y_test, y_predict)
print "sklearn SGDClassifier accuracy:"
print accuracy_score(y_test, y_preSGD)
print "my perceptron accuracy:"
print accuracy_score(y_test, y_pred)
print "\n"
#print clf.coef_
def x22(x1, w):
w0 = clf.coef_[0][0]
w1 = clf.coef_[0][1]
