0.理论
贝叶斯定理由英国数学家贝叶斯 ( Thomas Bayes 1702-1761 ) 发展,用来描述两个条件概率之间的关系,比如 P(A|B) 和 P(B|A)。按照乘法法则,可以立刻导出:P(A∩B) = P(A)*P(B|A)=P(B)*P(A|B)。如上公式也可变形为:P(A|B)=P(B|A)*P(A)/P(B)。
使用朴素贝叶斯对垃圾邮件进行分类, 假定各个单词之间没有联系。只是独立地检查邮件中的每个单词,然后将每个单词对确定是否为垃圾邮件的贡献组合起来。
1.用朴素贝叶斯实现垃圾邮件分类
import os
import io
import pandas as pd
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.naive_bayes import MultinomialNB
def resultFrameFromDirectory(root, classification):
"""读取文件夹路径下的所有文件数据,提取邮件内容和分类
Args:
root (str): 文件夹路径
classification (str): 分类
Returns:
resultFrame: 消息和分类
"""
files = os.listdir(root)
result = {'message': [], 'classification': []}
for file in files:
path = os.path.join(root, file)
inBody = False
lines = []
f = io.open(path, 'r', encoding='latin1')
for line in f:
if inBody:
lines.append(line)
elif line == '\n':
inBody = True
f.close()
message = '\n'.join(lines)
result['message'].append(message)
result['classification'].append(classification)
return pd.DataFrame(result)
root_ham = 'E:/python/python数据科学与机器学习/《Python数据科学与机器学习:从入门到实践》源代码/emails/ham'
root_spam = 'E:/python/python数据科学与机器学习/《Python数据科学与机器学习:从入门到实践》源代码/emails/spam'
df_ham = resultFrameFromDirectory(root_ham, 'ham')
df_spam = resultFrameFromDirectory(root_spam, 'spam')
df = df_ham.append(df_spam)
print(df.shape)
(3000, 2)
C:\Users\Administrator\AppData\Local\Temp\ipykernel_10964\3226765712.py:43: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.
df = df_ham.append(df_spam)
df_ham.head()
message | classification | |
---|---|---|
0 | Date: Wed, 21 Aug 2002 10:54:46 -05... | ham |
1 | Martin A posted:\n\nTassos Papadopoulos, the G... | ham |
2 | Man Threatens Explosion In Moscow \n\n\n\nThur... | ham |
3 | Klez: The Virus That Won't Die\n\n \n\nAlready... | ham |
4 | > in adding cream to spaghetti carbonara, whi... | ham |
df_spam.head()
message | classification | |
---|---|---|
0 | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Tr... | spam |
1 | 1) Fight The Risk of Cancer!\n\nhttp://www.adc... | spam |
2 | 1) Fight The Risk of Cancer!\n\nhttp://www.adc... | spam |
3 | ##############################################... | spam |
4 | I thought you might like these:\n\n1) Slim Dow... | spam |
vectorizer = CountVectorizer()
X = vectorizer.fit_transform(df['message'])
y = df['classification']
classifier = MultinomialNB()
classifier.fit(X, y)
# 预测
examples = ['Free Viagra now!!!', "Hi Bob, how about a game of golf tomorrow?"]
example_counts = vectorizer.transform(examples)
predictions = classifier.predict(example_counts)
predictions
array(['spam', 'ham'], dtype='<U4')
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
# 训练
classifier = MultinomialNB()
classifier.fit(X_train, y_train)
# 预测
y_pred = classifier.predict(X_test)
score = round(accuracy_score(y_test, y_pred)*100,2)
print('模型准确率:{}%'.format(score))
模型准确率:96.0%
2. 参考资料
Python数据科学与机器学习:从入门到实践
作者:
[美]弗兰克•凯恩(Frank Kane)