多标签（multilabel）分类

场景

• 一部电影可以同时属于爱情片和科幻片
• 一篇文章可以同时包含娱乐和汽车
• 一张图片可以同时出现猫和狗。
• 参考网址：https://scikit-learn.org/stable/modules/multiclass.html

多标签二值化

from sklearn.preprocessing import MultiLabelBinarizer
Y = [[], [1], [2], [1, 2]]
print(MultiLabelBinarizer().fit_transform(Y))

打印结果

[[0 0]
 [1 0]
 [0 1]
 [1 1]]

分类器

from sklearn.preprocessing import MultiLabelBinarizer
from sklearn.multiclass import OneVsRestClassifier
from sklearn.tree import DecisionTreeClassifier

# 样本
X = [
    [1, 0, 1, 0], [0, 1, 0, 1], [0, 1, 1, 0], [1, 0, 0, 1],
    [1, 1, 0, 0], [2, 0, 0, 0], [0, 2, 0, 0],
    [0, 0, 1, 1], [0, 0, 2, 0], [0, 0, 0, 2],
    [1, 1, 1, 1], [2, 0, 2, 0], [0, 2, 0, 2],
]
Y = [
    [], [], [], [],
    [1], [1], [1],
    [2], [2], [2],
    [1, 2], [1, 2], [1, 2],
]
X_test = [[1, 0, 0, 0], [0, 0, 0, 1], [3, 0, 3, 3]]
Y_test = [[], [], [1, 2]]

# 标签处理
transformer = MultiLabelBinarizer()
Y = transformer.fit_transform(Y)
Y_test = transformer.transform(Y_test)

# 建模
model = OneVsRestClassifier(DecisionTreeClassifier())
model.fit(X, Y)
print('训练得分：%.2f' % model.score(X, Y), '测试得分：%.2f' % model.score(X_test, Y_test))
print(model.predict(X_test))

打印结果

训练得分：1.00 测试得分：1.00
[[0 0]
 [0 0]
 [1 1]]

各种分类器比较

import warnings
warnings.filterwarnings('ignore')
from sklearn.preprocessing import MultiLabelBinarizer
from sklearn.multiclass import OneVsRestClassifier
from sklearn.naive_bayes import MultinomialNB
from sklearn.tree import DecisionTreeClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC
from sklearn.neural_network import MLPClassifier

# 样本
X = [
    [0, 0, 0, 0], [1, 0, 1, 0], [0, 1, 0, 1], [0, 1, 1, 0], [1, 0, 0, 1],
    [1, 1, 0, 0], [2, 0, 0, 0], [0, 2, 0, 0], [0, 0, 1, 1], [0, 0, 2, 0], [0, 0, 0, 2],
    [1, 1, 1, 1], [2, 0, 2, 0], [0, 2, 0, 2], [2, 0, 0, 2], [0, 2, 2, 0]
]
Y = [
    [], [], [], [], [],
    [1], [1], [1], [2], [2], [2],
    [1, 2], [1, 2], [1, 2], [1, 2], [1, 2]
]
X_test = [[1, 0, 0, 0], [0, 0, 0, 1], [3, 0, 3, 3], [0, 3, 0, 0]]
Y_test = [[], [], [1, 2], [1]]

# 多标签二值化
transformer = MultiLabelBinarizer()
Y = transformer.fit_transform(Y)
Y_test = transformer.transform(Y_test)

# 建模
for clf in [MultinomialNB, DecisionTreeClassifier, LogisticRegression, SVC, MLPClassifier]:
    model = OneVsRestClassifier(clf())
    model.fit(X, Y)
    print(clf.__name__, model.score(X, Y), model.score(X_test, Y_test))
    # print(model.predict(X_test))

打印结果

MultinomialNB 0.75 0.25
DecisionTreeClassifier 1.0 1.0
LogisticRegression 0.75 0.5
SVC 1.0 1.0
MLPClassifier 1.0 1.0

文本分类

from jieba import cut
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.preprocessing import MultiLabelBinarizer
from sklearn.multiclass import OneVsRestClassifier
from sklearn.linear_model import LogisticRegression

# 样本
X = '''
某女星代言小米手机
小米手机预售价2000元
某女星片酬2000万
月入2000元
'''.strip().split()
Y = [
    ['娱乐', '手机'],
    ['手机'],
    ['娱乐'],
    [],
]
X_test = '''
2000元
某女星买2000元的小米手机
'''.strip().split()
Y_test = [[], ['娱乐', '手机']]

# 文本向量化
vectorizer = TfidfVectorizer(tokenizer=cut)
X = vectorizer.fit_transform(X)
X_test = vectorizer.transform(X_test)

# 多标签二值化
transformer = MultiLabelBinarizer()
Y = transformer.fit_transform(Y)
Y_test = transformer.transform(Y_test)

# 建模
model = OneVsRestClassifier(LogisticRegression())
model.fit(X, Y)
print('训练得分：%.2f' % model.score(X, Y), '测试得分：%.2f' % model.score(X_test, Y_test))
print(model.predict_proba(X_test))

打印结果

训练得分：1.00 测试得分：1.00
[[0.39389651 0.45297039]
 [0.51392277 0.54656221]]