机器学习实战第四章-朴素贝叶斯

朴素贝叶斯

1.1准备数据：从文本中构建词向量

'''
创建从词表到向量的转换函数
'''
#创建实验样本
def loadDataSet():
    postingList=[['my', 'dog', 'has', 'flea', 'problems', 'help', 'please'],                #切分的词条
                 ['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代表正常言论
    return postingList,classVec

#创建所有文档中出现不重复词的列表
def createVocabList(dataSet):
    vocabSet=set([])
    for document in dataSet:
        vocabSet=vocabSet|set(document)   #取并集
    return list(vocabSet)                 #返回的是列表

#输出某个文档的向量，表示文档中词汇是否在词汇表中出现
def setOfWords2Vec(vocabList,inputSet):
    returnVec=[0]*len(vocabList)       #创建一个和词汇表等长的0向量
    for word in inputSet:
        if word in vocabList:
            returnVec[vocabList.index(word)]=1
        else:
            print('the word: %s is not in my vocabulary!'%word)
    return returnVec

#测试
listOPosts,listClasses=loadDataSet()
myVocabList=createVocabList(listOPosts)
myVocabList             #词汇表

['flea',
 'help',
 'is',
 'has',
 'I',
 'posting',
 'garbage',
 'maybe',
 'stop',
 'steak',
 'dog',
 'stupid',
 'not',
 'park',
 'ate',
 'quit',
 'mr',
 'licks',
 'food',
 'how',
 'him',
 'my',
 'so',
 'buying',
 'problems',
 'cute',
 'dalmation',
 'to',
 'love',
 'please',
 'worthless',
 'take']

setOfWords2Vec(myVocabList,listOPosts[0])

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

1.2 训练算法：从词向量计算概率

import numpy as np
def trainNB0(trainMatrix,trainCategory):
    numTrainDocs=len(trainMatrix)                   #训练文档数目
    numWords=len(trainMatrix[0])                    #每篇文档词条数
    pAbusive=sum(trainCategory)/float(numTrainDocs) #属于侮辱类文档的概率
    p0Num=np.zeros(numWords)                        #词条出现向量初始化为0
    p1Num=np.zeros(numWords)
    p0Denom=0.0                                     #分母
    p1Denom=0.0
    for i in range(numTrainDocs):
        if trainCategory[i]==1:
            p1Num+=trainMatrix[i]                   #向量相加
            p1Denom+=sum(trainMatrix[i])            #统计侮辱类条件概率p(w1|c1),p(w2|c1)……
        else:
            p0Num+=trainMatrix[i]                   #统计非侮辱类条件概率p(w1|c0),p(w2|c0)……
            p0Denom+=sum(trainMatrix[i])
    p1Vect=p1Num/p1Denom
    p0Vect=p0Num/p0Denom
    return p0Vect,p1Vect,pAbusive