朴素贝叶斯——文章分类示例

朴素贝叶斯——文章分类示例

0. 原理-公式推导

• Venn图
  
• 因为
  $$P(A|B)=\frac{P(A\bigcap B)}{P(B)}=>P(A\bigcap B)=P(A|B)\cdot P(B)$$
  $$P(B|A)=\frac{P(A\bigcap B)}{P(A)}=>P(A\bigcap B)=P(B|A)\cdot P(A)$$
  $$P(A|B)\cdot P(B)=P(B|A)\cdot P(A)$$
• 所以
  $$贝叶斯公式:P(A|B)=\frac{P(B|A)\cdot P(A)}{P(B)}$$

1. 导包

import pandas as pd
from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.naive_bayes import MultinomialNB

2. 原始数据

# fetch_20newsgroups新闻数据下载需要一定时间
news = datasets.fetch_20newsgroups(subset="all")
X = news.data
y = news.target
print(X[0:5])
print(y[0:5])

# 数据示例
["From: Mamatha Devineni Ratnam <mr47+@andrew.cmu.edu>\nSubject: Pens fans reactions\nOrganization: Post Office, Carnegie Mellon,  ...... I was very disappointed not to see the Islanders lose the final\nregular season game.          PENS RULE!!!\n\n"]
# 新闻分类示例
[10  3 17  3  4]

3. 数据预处理

# 训练集、测试集拆分
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25)
# 使用TF-IDF
tf_idf = TfidfVectorizer()
X_train = tf_idf.fit_transform(X_train)
X_test = tf_idf.transform(X_test)
print(X_train[0:5])
print(tf_idf.get_feature_names()[8000:8020]) # 打印部分分词特征名

4. 构建朴素贝叶斯模型

mlt = MultinomialNB(alpha=1.0)
mlt.fit(X_train, y_train)

5. 结果预测与评分

# 预测
y_predict = mlt.predict(X_test)
print("预测结果", y_predict[0:10])
print("实际结果", y_test[0:10])
# 评分
print("准确度", mlt.score(X_test, y_test))

# 结果示例
预测结果 [ 0 16 15 12 16 12  6 17  7 11]
实际结果 [19 16 15 12 16 12  6 17  7 11]
准确度 0.8529286926994907