机器学习实战-朴素贝叶斯

对于事件A和B，贝叶斯定理的表达式可写成P(A|B) = P(A)P(B|A) / P(B)

P(A)   ：先验概率

P(A|B) ：后验概率

P(B|A)  ：似然度

 P(B) ：标准化常量

朴素贝叶斯机器学习实战：

#encoding:utf-8
from numpy import *
#下载数据
#贝努力模型~多项式模型
#贝努力~~不考虑词出现的次数~~只考虑出不出现
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表示正常言论
    return postingList,classVec
#不重复词列表                 
def createVocabList(dataSet):
    vocabSet = set([])  #创建一个空集~~set类型数据~返回的是不重复词的列表
    for document in dataSet:
        vocabSet = vocabSet | set(document) #创建两个集合的并集
    return list(vocabSet)
#输入：文档，不重复词列表   输出：文档向量~~向量每一元素为1或0~~分别表示单词在文档中是否出现
#即：将单词转换为数字
def setOfWords2Vec(vocabList, inputSet):
    returnVec = [0]*len(vocabList)  #创建一个所含元素都为0的向量
    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
#输入：文档矩阵、类别向量   输出：正常词概率、侮辱性词概率、属于侮辱性文档概率
def trainNB0(trainMatrix,trainCategory):
    #计算class=1的概率p(1)~因为是二分类~因此class=0的概率为p(0)=1-p(1)
    numTrainDocs = len(trainMatrix)
    numWords = len(trainMatrix[0])
    pAbusive = sum(trainCategory)/float(numTrainDocs)
    p0Num = ones(numWords); p1Num = ones(numWords)      #change to ones() 
    p0Denom = 2.0; p1Denom = 2.0                        #change to 2.0~初始化概率~~值均为0是作用于二分类，值均为2可以直接应用于多类别分类
    for i in range(numTrainDocs):
        if trainCategory[i] == 1:#出现侮辱性文字
            p1Num += trainMatrix[i]#该侮辱性词个数+1
            p1Denom += sum(trainMatrix[i]) #文档侮辱性总词数+1
        else:
            p0Num += trainMatrix[i]#正常词数
            p0Denom += sum(trainMatrix[i])
    p1Vect = log(p1Num/p1Denom)          #change to log()
    p0Vect = log(p0Num/p0Denom)          #change to log()#对每个元素作除法
    return p0Vect,p1Vect,pAbusive
#贝叶斯分类函数
#输入：要分类的向量vec2Classify，以及trainNB0得到的3个概率   输出：得到类别
def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):
    p1 = sum(vec2Classify * p1Vec) + log(pClass1)    #element-wise mult~元素相乘
    p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)#元素相乘
    if p1 > p0:
        return 1
    else: 
        return 0
#输入：文档、词列表   输出：文档向量~~向量的值表示单词在文档中出现的次数
#即：将单词转换为数字  
def bagOfWords2VecMN(vocabList, inputSet):
    returnVec = [0]*len(vocabList)
    for word in inputSet:
        if word in vocabList:
            returnVec[vocabList.index(word)] += 1
    return returnVec
#贝叶斯测试~~应用：过滤网站恶意留言
def testingNB():
    listOPosts,listClasses = loadDataSet()#下载数据和类别
    myVocabList = createVocabList(listOPosts)#建立不重复词列表
    trainMat=[]#训练数据
    for postinDoc in listOPosts:
        trainMat.append(setOfWords2Vec(myVocabList, postinDoc))#词转换成向量
    p0V,p1V,pAb = trainNB0(array(trainMat),array(listClasses))#正常词概率、侮辱性词概率、属于侮辱性文档概率
    testEntry = ['love', 'my', 'dalmation']#测试数据
    thisDoc = array(setOfWords2Vec(myVocabList, testEntry))#测试数据转向量
    print testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb)#测试结果
    testEntry = ['stupid', 'garbage']
    thisDoc = array(setOfWords2Vec(myVocabList, testEntry))
    print testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb)
#文件解析。email
#输入：比较大的字符串数据   输出：单词列表
def textParse(bigString):    #input is big string, #output is word list
    import re
    listOfTokens = re.split(r'\W*', bigString)#正则匹配、字符串分割
    return [tok.lower() for tok in listOfTokens if len(tok) > 2] #1.将所有字符串转为小写，2.找出字符串长度大于2的
#垃圾邮件测试~得到分类错误率 email
def spamTest():
    docList=[]; classList = []; fullText =[]
    #导入并解析文本文件
    for i in range(1,26):
        wordList = textParse(open('email/spam/%d.txt' % i).read())
        docList.append(wordList)
        fullText.extend(wordList)
        classList.append(1)
        wordList = textParse(open('email/ham/%d.txt' % i).read())
        docList.append(wordList)
        fullText.extend(wordList)
        classList.append(0)
    vocabList = createVocabList(docList)#创建无重复词列表
    trainingSet = range(50); testSet=[]           #create test set
    for i in range(10):#随机选取10个为测试集
        randIndex = int(random.uniform(0,len(trainingSet)))
        testSet.append(trainingSet[randIndex])
        del(trainingSet[randIndex])  
    trainMat=[]; trainClasses = []
    for docIndex in trainingSet:#train the classifier (get probs) trainNB0
        trainMat.append(bagOfWords2VecMN(vocabList, docList[docIndex]))
        trainClasses.append(classList[docIndex])
    p0V,p1V,pSpam = trainNB0(array(trainMat),array(trainClasses))
    errorCount = 0
    #对测试集分类
    for docIndex in testSet:        #classify the remaining items
        wordVector = bagOfWords2VecMN(vocabList, docList[docIndex])
        if classifyNB(array(wordVector),p0V,p1V,pSpam) != classList[docIndex]:
            errorCount += 1
            print "classification error",docList[docIndex]
    print 'the error rate is: ',float(errorCount)/len(testSet)#分类错误的占比

测试代码：

#贝叶斯测试~~应用：过滤网站恶意留言
bayes.testingNB()
#垃圾邮件测试~得到分类错误率 email
bayes.spamTest()
bayes.spamTest()#输出错误率~以及哪些文档出错