朴素贝叶斯

今天看来朴素贝叶斯，最大的进步我觉得就是知道了她的一个非常重要的用途，用作文本分类任务，下面是copy两个大佬的代码，用来进行记忆：

#!/usr/bin/python
# coding:utf-8
from numpy import *
from math import *


def loadDataSet():
    postingList = [['my', 'dog', 'has', 'flea', 'problems', 'help', 'please'],
                   ['maybe', 'not', 'take', 'him', 'to', 'dog', 'park', 'stupid'],
                   ['my', 'dalmation', 'is', 'so', 'cute', 'I', 'love', 'him'],
                   ['stop', 'posting', 'stupid', 'worthless', 'garbage'],
                   ['mr', 'licks', 'ate', 'my', 'steak',
                       'how', 'to', 'stop', 'him'],
                   ['quit', 'buying', 'worthless', 'dog', 'food', 'stupid']]
    classVec = [0, 1, 0, 1, 0, 1]  # 1代表侮辱性文字 0代表正常的言论
    return postingList, classVec


def createVocabList(dataSet):    #  给定一个数据集，创建一个不含重复单词的词向量
    vocabList = set([])
    for document in dataSet:
        vocabList = vocabList | set(document)
    return list(vocabList)


def setOfWords2Vec(vocabList, inputSet):   #给定一个词向量和一条单词构成的记录，生成次条记录的向量
    returnVec = [0] * len(vocabList)
    for word in inputSet:
        if word in vocabList:
            returnVec[vocabList.index(word)] = 1
    return returnVec


def trainNB0(trainMatrix, trainCategory):  # 训练参数,得到一个参数矩阵，对应着各个单词对应分类的出现频率
    numTrainDocs = len(trainMatrix)
    numWords = len(trainMatrix[0])
    pAbusive = sum(trainCategory) / float(numTrainDocs)
    p0Num = ones(numWords)
    p1Num = ones(numWords)
    p0Denom = 2
    p1Denom = 2
    for i in range(numTrainDocs):
        if trainCategory[i] == 1:
            p1Num += trainMatrix[i]
            p1Denom += sum(trainMatrix[i])
        else:
            p0Num += trainMatrix[i]
            p0Denom += sum(trainMatrix[i])
    p1Vec = p1Num / p1Denom
    p0Vec = p0Num / p0Denom
    return p0Vec, p1Vec, pAbusive


def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):   #  计算分类
    p1 = sum(vec2Classify * p1Vec) + log(pClass1)  # array相乘并且加上类别个数的对数上
    p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)
    if p1 > p0:
        return 1
    else:
        return 0


def testNB():    # 测试贝叶斯
    listOPosts, listClasses = loadDataSet()   #加载数据集
    myVocabList = createVocabList(listOPosts)  # 得到词向量
    trainMat = []
    for postinDoc in listOPosts:  # 生成训练数据集的向量
        vecOfWord = setOfWords2Vec(myVocabList, postinDoc)
        trainMat.append(vecOfWord)
    p0v, p1v, pAb = trainNB0(trainMat, listClasses)  # 根据已有的数据训练出的参数p0 p1为矩阵
    testEntry = ['stupid']  # 测试数据
    testMatrix = array(setOfWords2Vec(myVocabList, testEntry)) # 得到测试数据的array
    return classifyNB(testMatrix, p0v, p1v, pAb)  # 分类


print(testNB()) #测试主函数

下面是调用sklearn实现的，用的语料库是fetch_20newsgroups中的新闻：

#coding=utf-8

from sklearn.datasets import fetch_20newsgroups  # 从sklearn.datasets里导入新闻数据抓取器 fetch_20newsgroups
from sklearn.model_selection import  train_test_split
from sklearn.feature_extraction.text import CountVectorizer  # 从sklearn.feature_extraction.text里导入文本特征向量化模块
from sklearn.naive_bayes import MultinomialNB     # 从sklean.naive_bayes里导入朴素贝叶斯模型
from sklearn.metrics import classification_report

#1.数据获取
news = fetch_20newsgroups(subset='all')
print (len(news.data)) # 输出数据的条数：18846

#2.数据预处理：训练集和测试集分割，文本特征向量化
X_train,X_test,y_train,y_test = train_test_split(news.data,news.target,test_size=0.25,random_state=33) # 随机采样25%的数据样本作为测试集

#print y_train[0:100]  #查看标签

#文本特征向量化
vec = CountVectorizer()
X_train = vec.fit_transform(X_train)
X_test = vec.transform(X_test)
print(X_train[0],len(X_train)) #查看训练样本,估计有很多维，并打印出维数
#3.使用朴素贝叶斯进行训练
mnb = MultinomialNB()   # 使用默认配置初始化朴素贝叶斯
mnb.fit(X_train,y_train)    # 利用训练数据对模型参数进行估计
y_predict = mnb.predict(X_test)     # 对参数进行预测

#4.获取结果报告

#验证正确率
from sklearn.metrics import accuracy_score
print('Accuracy : % .2f' % accuracy_score(y_predict, y_test))

print ('The Accuracy of Naive Bayes Classifier is:', mnb.score(X_test,y_test))
print (classification_report(y_test, y_predict, target_names = news.target_names))

我觉得最重要的不是实现，而是明白它的工作原理，明白原理之后你才能自己去写出比较好的，适用的代码