本文整理汇总了Python中sklearn.naive_bayes.MultinomialNB.set_params方法的典型用法代码示例。如果您正苦于以下问题：Python MultinomialNB.set_params方法的具体用法？Python MultinomialNB.set_params怎么用？Python MultinomialNB.set_params使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类sklearn.naive_bayes.MultinomialNB的用法示例。
在下文中一共展示了MultinomialNB.set_params方法的5个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: sk_learn_NB
# 需要导入模块: from sklearn.naive_bayes import MultinomialNB [as 别名]
# 或者: from sklearn.naive_bayes.MultinomialNB import set_params [as 别名]
def sk_learn_NB(a, jane_data, sherlock_data, k, e, l) :
NB = MultinomialNB()
NB.set_params(alpha=a)
#Combine the data for each snippet into one data list and create a target list with corresponding class labels
data = []
target = []
for d in jane_data :
data.append(d)
target.append(0)
for d in sherlock_data :
data.append(d)
target.append(1)
data_array = numpy.array(data)
target_array = numpy.array(target)
#K-fold cross-validation
kf = cross_validation.KFold(len(data), n_folds = k, shuffle = True)
fold_stats = []
for train_index, test_index in kf:
data_train, data_test = data_array[train_index], data_array[test_index]
target_train, target_test = target_array[train_index], target_array[test_index]
NB.fit(data_train, target_train)
find_error(NB.predict(data_test), target_test, fold_stats)
#Determining average stats for k-fold cross-validation
recall_overall_avg = 0.0
precision_overall_avg = 0.0
fstat_overall_avg = 0.0
recall_list = []
precision_list = []
fstat_list = []
for i in range(0,len(fold_stats)) :
recall_overall_avg = recall_overall_avg + fold_stats[i][0]
precision_overall_avg = precision_overall_avg + fold_stats[i][1]
fstat_overall_avg = fstat_overall_avg + fold_stats[i][2]
recall_list.append(fold_stats[i][0])
precision_list.append(fold_stats[i][1])
fstat_list.append(fold_stats[i][2])
recall_overall_avg = recall_overall_avg/len(fold_stats)
precision_overall_avg = precision_overall_avg/len(fold_stats)
fstat_overall_avg = fstat_overall_avg/len(fold_stats)
var_recall = numpy.var(recall_list)
var_precision = numpy.var(precision_list)
var_fstat = numpy.var(fstat_list)
print("Average stats for", k, "-fold cross-validation with l=", l, "and e=", e)
print("Recall: ", recall_overall_avg, "with variance:", var_recall)
print("Precision: ", precision_overall_avg, "with variance:", var_precision)
print("F-statistic: ", fstat_overall_avg, "with variance:", var_fstat)示例2: alpha_search
# 需要导入模块: from sklearn.naive_bayes import MultinomialNB [as 别名]
# 或者: from sklearn.naive_bayes.MultinomialNB import set_params [as 别名]
def alpha_search(vectorizer):
dg = DataGatherer()
y_test = dg.validate_target
y_train = dg.labeled_target
X_train = vectorizer.fit_transform(dg.labeled_data)
X_test = vectorizer.transform(dg.validate_data)
nb = MultinomialNB()
alpha = 0.0
best_score = -1
best_alpha = None
while(alpha < 5):
alpha += 0.5
nb.set_params(alpha = alpha)
nb.fit(X_train, y_train)
score = nb.score(X_test, y_test)
if score > best_score:
best_score = score
best_alpha = alpha
print(str(alpha) + "," + str(score))
print("")
print("Best alph: " + str(best_alpha))
print("Best score: " + str(best_score))示例3: __init__
# 需要导入模块: from sklearn.naive_bayes import MultinomialNB [as 别名]
# 或者: from sklearn.naive_bayes.MultinomialNB import set_params [as 别名]
class SparseModel:
model = None
model_type = None
default = None
column_names = None
class_index = None
params = None
trained = None
w = None
b = None
metric = None
def __init__(self, model_type=None, column_names=None, metric='f1', **kwargs):
self.model_type = model_type
self.column_names = column_names
self.params = kwargs
self.trained = None
self.metric = metric
if model_type == 'LR':
if self.params.get('regularization', None) is None:
self.params['regularization'] = 'l1'
if self.params.get('alpha', None) is None:
self.params['alpha'] = 1.0
self.model = lr(penalty=self.params['regularization'], C=self.params['alpha'])
elif model_type == 'SVM' or model_type == 'SVMNB':
if self.params.get('kernel', None) is None:
self.params['kernel'] = 'rbf'
if model_type == 'SVM':
if self.params.get('alpha', None) is None:
self.params['alpha'] = 0.1
else: # elif model_type == SVMNB:
self.params['kernel'] = 'linear'
if self.params.get('alpha', None) is None:
self.params['alpha'] = 1
if self.params.get('beta', None) is None:
self.params['beta'] = 0.25
if self.params['kernel'] == 'linear':
# override regularization parameter to avoid a conflict
self.params['regularization'] = 'l2'
self.model = svm.LinearSVC(C=self.params['alpha'])
else: # elif self.params['kernel'] != 'linear':
if self.params.get('degree', None) is None:
self.params['degree'] = 3
if self.params.get('gamma', None) is None:
self.params['gamma'] = 0.0
if self.params.get('coef0', None) is None:
self.params['coef0'] = 0.0
self.model = svm.SVC(C=self.params['alpha'], kernel=self.params['kernel'], degree=self.params['degree'],
gamma=self.params['gamma'], coef0=self.params['coef0'])
elif model_type == 'MNB':
if 'alpha' not in self.params:
self.params['alpha'] = 1.0
self.model = MultinomialNB(alpha=self.params['alpha'], fit_prior=True)
elif model_type == 'myMNB':
if 'alpha' not in self.params:
self.params['alpha'] = 1.0
self.model = None
else:
self.model_type = 'default'
self.model = None
def set_alpha(self, alpha):
self.params['alpha'] = alpha
if self.model_type == 'LR':
self.model.set_params(C=alpha)
elif self.model_type == 'SVM' or self.model_type == 'SVMNB':
self.model.set_params(C=alpha)
elif self.model_type == 'MNB':
self.model.set_params(alpha=float(alpha))
def fit(self, X, y, sample_weights=None):
n_items, n_features = X.shape
if y.sum() == 0 or y.sum() == n_items or self.model_type == 'default':
counts = np.bincount(y)
y_mode = np.argmax(counts)
self.default = y_mode
self.model = None
self.model_type = 'default'
elif self.model_type == 'myMNB':
# assumes y = {0,1}
assert len(np.bincount(y)) == 2
index_0 = np.array(y) == 0
index_1 = np.array(y) == 1
p = np.sum(X.toarray()[index_1, :], axis=0) + float(self.params['alpha'])
q = np.sum(X.toarray()[index_0, :], axis=0) + float(self.params['alpha'])
self.w = np.log((p / float(np.sum(np.abs(p)))) / (q / float(np.sum(np.abs(q)))))
self.b = np.log(np.sum(index_1) / float(np.sum(index_0)))
if np.isnan(self.w).any():
sys.exit("nans encountered in " + str(self.__class__) + "." + self.fit.__name__)
elif self.model_type == 'SVMNB':
# assumes y = {0,1}
assert len(np.bincount(y)) == 2
# first calculate the inputs we would use for MNB
X_dense = X.toarray()
_, n_features = X_dense.shape
index_0 = np.array(y) == 0
index_1 = np.array(y) == 1
#.........这里部分代码省略.........
示例4: __init__
# 需要导入模块: from sklearn.naive_bayes import MultinomialNB [as 别名]
# 或者: from sklearn.naive_bayes.MultinomialNB import set_params [as 别名]
class SparseModel:
model = None
model_type = None
default = None
column_names = None
params = None
trained = None
w = None
b = None
def __init__(self, model_type=None, column_names=None, **kwargs):
self.model_type = model_type
self.column_names = column_names
self.params = kwargs
self.trained = None
if model_type == 'LR':
if self.params.get('regularization', None) is None:
self.params['regularization'] = 'l1'
if self.params.get('alpha', None) is None:
self.params['alpha'] = 1.0
self.model = lr(penalty=self.params['regularization'], C=self.params['alpha'])
elif model_type == 'SVM' or model_type == 'SVMNB':
if self.params.get('kernel', None) is None:
self.params['kernel'] = 'rbf'
if model_type == 'SVM':
if self.params.get('alpha', None) is None:
self.params['alpha'] = 0.1
else: # elif model_type == SVMNB:
self.params['kernel'] = 'linear'
if self.params.get('alpha', None) is None:
self.params['alpha'] = 1
if self.params.get('beta', None) is None:
self.params['beta'] = 0.25
if self.params['kernel'] == 'linear':
# override regularization parameter to avoid a conflict
self.params['regularization'] = 'l2'
self.model = svm.LinearSVC(C=self.params['alpha'])
else: # elif self.params['kernel'] != 'linear':
if self.params.get('degree', None) is None:
self.params['degree'] = 3
if self.params.get('gamma', None) is None:
self.params['gamma'] = 0.0
if self.params.get('coef0', None) is None:
self.params['coef0'] = 0.0
self.model = svm.SVC(C=self.params['alpha'], kernel=self.params['kernel'], degree=self.params['degree'],
gamma=self.params['gamma'], coef0=self.params['coef0'])
elif model_type == 'MNB':
if 'alpha' not in self.params:
self.params['alpha'] = 1.0
self.model = MultinomialNB(alpha=self.params['alpha'], fit_prior=True)
elif model_type == 'myMNB':
if 'alpha' not in self.params:
self.params['alpha'] = 1.0
self.model = None
else:
self.model_type = 'default'
self.model = None
def set_alpha(self, alpha):
self.params['alpha'] = alpha
if self.model_type == 'LR':
self.model.set_params(C=alpha)
elif self.model_type == 'SVM' or self.model_type == 'SVMNB':
self.model.set_params(C=alpha)
elif self.model_type == 'MNB':
self.model.set_params(alpha=float(alpha))
def fit(self, X, y):
n_items, n_features = X.shape
if y.sum() == 0 or y.sum() == n_items or self.model_type == 'default':
counts = np.bincount(y)
y_mode = np.argmax(counts)
self.default = y_mode
self.model = None
self.model_type = 'default'
elif self.model_type == 'myMNB':
# assumes y = {0,1}
assert len(np.bincount(y)) == 2
index_0 = np.array(y) == 0
index_1 = np.array(y) == 1
p = np.sum(X.toarray()[index_1, :], axis=0) + float(self.params['alpha'])
q = np.sum(X.toarray()[index_0, :], axis=0) + float(self.params['alpha'])
self.w = np.log((p / float(np.sum(np.abs(p)))) / (q / float(np.sum(np.abs(q)))))
self.b = np.log(np.sum(index_1) / float(np.sum(index_0)))
if np.isnan(self.w).any():
sys.exit("nans encountered in " + str(self.__class__) + "." + self.fit.__name__)
elif self.model_type == 'SVMNB':
# assumes y = {0,1}
assert len(np.bincount(y)) == 2
# first calculate the inputs we would use for MNB
X_dense = X.toarray()
_, n_features = X_dense.shape
index_0 = np.array(y) == 0
index_1 = np.array(y) == 1
# assume a hyperparameter for MNB of 1.0
p = np.sum(X_dense[index_1, :], axis=0) + 1
q = np.sum(X_dense[index_0, :], axis=0) + 1
r = np.log((p / float(np.sum(np.abs(p)))) / (q / float(np.sum(np.abs(q)))))
#.........这里部分代码省略.........
示例5: __init__
# 需要导入模块: from sklearn.naive_bayes import MultinomialNB [as 别名]
# 或者: from sklearn.naive_bayes.MultinomialNB import set_params [as 别名]
class NaiveBayes:
def __init__(self, alpha):
self.classifier = MultinomialNB(alpha=alpha, fit_prior=False)
self.creation_ts = datetime.now()
def train(self, training_set):
features = []
classes = []
for index, entry in training_set.iterrows():
features.append(entry[['Osfamily', 'Publisher', 'ArticleDistance']])
# This is for when article popularity is known
# features.append(entry[['UserClicksAd', 'Osfamily', 'Publisher', 'ArticleDistance', 'ArticlePopularity']])
classes.append(entry['Output'])
print('... Training the model')
self.classifier.fit(features, classes)
def classification(self, train_data):
features = []
for index, entry in train_data.iterrows():
features.append(entry[['Osfamily', 'Publisher', 'ArticleDistance']])
# This is for when article popularity is known
# features.append(entry[['UserClicksAd', 'Osfamily', 'Publisher', 'ArticleDistance', 'ArticlePopularity']])
print('... Started classification')
results_prob = self.classifier.predict(features)
return results_prob
def predict_prob(self, train_data):
features = []
for index, entry in train_data.iterrows():
features.append(entry[['Osfamily', 'Publisher', 'ArticleDistance']])
# This is for when article popularity is known
# features.append(entry[['UserClicksAd', 'Osfamily', 'Publisher', 'ArticleDistance', 'ArticlePopularity']])
print('... Started classification')
results_prob = self.classifier.predict_proba(features)
return results_prob
def series_predict_prob(self, data):
features = []
features.append(data[['Osfamily', 'Publisher', 'ArticleDistance']])
# This is for when article popularity is known
# features.append(entry[['UserClicksAd', 'Osfamily', 'Publisher', 'ArticleDistance', 'ArticlePopularity']])
results_prob = self.classifier.predict_proba(features)
return results_prob
def classification_CV(self, train_data, number_of_folds):
features = []
for index, entry in train_data.iterrows():
features.append(entry[['Osfamily', 'Publisher', 'ArticleDistance']])
# This is for when article popularity is known
# features.append(entry[['UserClicksAd', 'Osfamily', 'Publisher', 'ArticleDistance', 'ArticlePopularity']])
print('... Started crossfold validation')
score = cross_validation.cross_val_score(self.classifier, features, train_data.Output, cv=number_of_folds,
scoring='f1')
return score
def set_priors(self, priors):
# sets priors of naive bayes component to given values
if priors[0] + priors[1] == 1:
self.classifier.set_params(class_prior=priors)
else:
print 'Priors should some up to 1.'
exit()