def classify():
if not os.path.exists(dp.android_dataset_dir):
print('数据预处理并保存中')
android_dataset = dp.load_file_to_dataset()
dp.save_android_dataset(android_dataset, dp.android_dataset_dir)
else:
android_dataset = dp.load_android_dataset(dp.android_dataset_dir)
train_start_time = time.perf_counter()
svm_clf = svm.SVC()
svm_clf.fit(android_dataset.train_features,
android_dataset.train_class_labels)
train_end_time = time.perf_counter()
test_features = android_dataset.test_features
test_class_labels = android_dataset.test_class_labels
test_num = android_dataset.test_num
predict_start_time = time.perf_counter()
acc = 0
for i in range(test_num):
class_label = svm_clf.predict([test_features[i]])[0]
if class_label == test_class_labels[i]:
acc += 1
print('正确', class_label)
else:
print('错误, 预测类别：%d, 正确类别：%d' % (class_label, test_class_labels[i]))
print('\n正确率：%.2f%%' % (100.0 * acc / test_num))
predict_end_time = time.perf_counter()
print('SVM-sklearn-训练阶段运行时间：%s秒' % (train_end_time - train_start_time))
print('SVM-sklearn-预测阶段运行时间：%s秒' % (predict_end_time - predict_start_time))
def classify():
if not os.path.exists(dp.android_dataset_dir):
print('数据预处理并保存中')
android_dataset = dp.load_file_to_dataset()
dp.save_android_dataset(android_dataset, dp.android_dataset_dir)
else:
android_dataset = dp.load_android_dataset(dp.android_dataset_dir)
test_features = android_dataset.test_features
test_class_labels = android_dataset.test_class_labels
test_num = android_dataset.test_num
train_features = android_dataset.train_features
train_class_labels = android_dataset.train_class_labels
predict_start_time = time.perf_counter()
acc = 0
for i in range(test_num):
class_label = knn.classify(test_features[i], train_features,
train_class_labels, 7)
if class_label == test_class_labels[i]:
acc += 1
print('正确', class_label)
else:
print('错误, 预测类别：%d, 正确类别：%d' % (class_label, test_class_labels[i]))
print('\n正确率：%.2f%%' % (100.0 * acc / test_num))
predict_end_time = time.perf_counter()
print('KNN-预测阶段运行时间：%s秒' % (predict_end_time - predict_start_time))
def classify():
if not os.path.exists(dp.android_dataset_dir):
print('数据预处理并保存中')
android_dataset = dp.load_file_to_dataset()
dp.save_android_dataset(android_dataset, dp.android_dataset_dir)
else:
android_dataset = dp.load_android_dataset(dp.android_dataset_dir)
train_start_time = time.perf_counter()
p_words, p_class = nb.train_nb(android_dataset.train_features,
android_dataset.train_class_labels,
android_dataset.class_num)
train_end_time = time.perf_counter()
test_features = android_dataset.test_features
test_class_labels = android_dataset.test_class_labels
test_num = android_dataset.test_num
class_num = android_dataset.class_num
predict_start_time = time.perf_counter()
acc = 0
for i in range(test_num):
class_label = nb.nb_classify(p_words, p_class, test_features[i],
class_num)
if class_label == test_class_labels[i]:
acc += 1
print('正确', class_label)
else:
print('错误, 预测类别：%d, 正确类别：%d' % (class_label, test_class_labels[i]))
print('\n正确率：%.2f%%' % (100.0 * acc / test_num))
predict_end_time = time.perf_counter()
print('naive-bayes-训练阶段运行时间：%s秒' % (train_end_time - train_start_time))
print('naive-bayes-预测阶段运行时间：%s秒' % (predict_end_time - predict_start_time))
def classify():
build_tree_start_time = 0
build_tree_end_time = 0
if os.path.exists(tree_file_dir):
print('决策树已保存')
d_tree, feature_labels, train_data_set, test_features_set, test_class_labels = id3.load_tree(
tree_file_dir)
else:
print('决策树未保存，重新建树中')
if not os.path.exists(dp.android_dataset_dir):
print('数据预处理并保存中')
android_dataset = dp.load_file_save_dataset()
else:
android_dataset = dp.load_android_dataset(dp.android_dataset_dir)
train_data_set = android_dataset.get_combined_train()
feature_labels = android_dataset.feature_labels
test_features_set = android_dataset.test_features
test_class_labels = android_dataset.test_class_labels
# print(feature_labels)
build_tree_start_time = time.perf_counter()
d_tree = id3.create_tree(train_data_set, feature_labels)
build_tree_end_time = time.perf_counter()
id3.store_tree(tree_file_dir, d_tree, feature_labels, train_data_set,
test_features_set, test_class_labels)
print(d_tree)
predict_start_time = time.perf_counter()
acc = 0
for i in range(len(test_features_set)):
class_label = id3.classify(d_tree, feature_labels,
test_features_set[i])
if class_label == test_class_labels[i]:
acc += 1
print('正确', class_label)
else:
print('错误, 预测类别：%d, 正确类别：%d' % (class_label, test_class_labels[i]))
print('\n正确率：%.2f%%' % (100.0 * acc / len(test_features_set)))
predict_end_time = time.perf_counter()
print('决策树-构建决策树运行时间：%s秒' % (build_tree_end_time - build_tree_start_time))
print('决策树-预测阶段运行时间：%s秒' % (predict_end_time - predict_start_time))
def classify():
if not os.path.exists(dp.android_dataset_dir):
print('数据预处理并保存中')
android_dataset = dp.load_file_to_dataset()
dp.save_android_dataset(android_dataset, dp.android_dataset_dir)
else:
android_dataset = dp.load_android_dataset(dp.android_dataset_dir)
test_features = android_dataset.test_features
test_class_labels = android_dataset.test_class_labels
test_num = android_dataset.test_num
train_start_time = time.perf_counter()
clf = MLPClassifier(hidden_layer_sizes=(100, ),
solver='adam',
activation='relu',
random_state=1,
verbose=True)
clf.fit(android_dataset.train_features, android_dataset.train_class_labels)
train_end_time = time.perf_counter()
predict_start_time = time.perf_counter()
predict_class_labels = clf.predict(test_features)
acc = 0
for i in range(len(predict_class_labels)):
class_label = predict_class_labels[i]
if class_label == test_class_labels[i]:
acc += 1
print('正确', class_label)
else:
print('错误, 预测类别：%d, 正确类别：%d' % (class_label, test_class_labels[i]))
print('\n正确率：%.2f%%' % (100.0 * acc / test_num))
predict_end_time = time.perf_counter()
print('Neural Network-训练阶段运行时间：%s秒' % (train_end_time - train_start_time))
print('Neural Network-预测阶段运行时间：%s秒' %
(predict_end_time - predict_start_time))
def classify():
if not os.path.exists(dp.android_dataset_dir):
print('数据预处理并保存中')
android_dataset = dp.load_file_to_dataset()
dp.save_android_dataset(android_dataset, dp.android_dataset_dir)
else:
android_dataset = dp.load_android_dataset(dp.android_dataset_dir)
train_start_time = time.perf_counter()
data = np.array(android_dataset.train_features)
label = np.array(class_labels_process(android_dataset.train_class_labels))
smo = svm.PlattSMO(data, label, 1, 0.0001, 10000, name='rbf', theta=20)
smo.smoP()
train_end_time = time.perf_counter()
test_features = android_dataset.test_features
test_class_labels = android_dataset.test_class_labels
test_num = android_dataset.test_num
predict_start_time = time.perf_counter()
predict_labels = smo.predict(test_features)
acc = 0
for i in range(len(predict_labels)):
class_label = predict_labels[i]
if class_label == test_class_labels[i]:
acc += 1
print('正确', class_label)
else:
print('错误, 预测类别：%d, 正确类别：%d' % (class_label, test_class_labels[i]))
print('\n正确率：%.2f%%' % (100.0 * acc / test_num))
predict_end_time = time.perf_counter()
print('SVM-训练阶段运行时间：%s秒' % (train_end_time - train_start_time))
print('SVM-预测阶段运行时间：%s秒' % (predict_end_time - predict_start_time))