樸素貝葉斯

算法優(yōu)缺點(diǎn)

• 優(yōu)點(diǎn)：在數(shù)據(jù)較少的情況下依然有效，可以處理多類別問(wèn)題

• 缺點(diǎn)：對(duì)輸入數(shù)據(jù)的準(zhǔn)備方式敏感

• 適用數(shù)據(jù)類型：標(biāo)稱型數(shù)據(jù)

算法思想：

樸素貝葉斯
比如我們想判斷一個(gè)郵件是不是垃圾郵件，那么我們知道的是這個(gè)郵件中的詞的分布，那么我們還要知道：垃圾郵件中某些詞的出現(xiàn)是多少，就可以利用貝葉斯定理得到。
樸素貝葉斯分類器中的一個(gè)假設(shè)是：每個(gè)特征同等重要

函數(shù)

loadDataSet()

創(chuàng)建數(shù)據(jù)集，這里的數(shù)據(jù)集是已經(jīng)拆分好的單詞組成的句子，表示的是某論壇的用戶評(píng)論，標(biāo)簽1表示這個(gè)是罵人的

createVocabList(dataSet)

找出這些句子中總共有多少單詞，以確定我們?cè)~向量的大小

setOfWords2Vec(vocabList, inputSet)

將句子根據(jù)其中的單詞轉(zhuǎn)成向量，這里用的是伯努利模型，即只考慮這個(gè)單詞是否存在

bagOfWords2VecMN(vocabList, inputSet)

這個(gè)是將句子轉(zhuǎn)成向量的另一種模型，多項(xiàng)式模型，考慮某個(gè)詞的出現(xiàn)次數(shù)

trainNB0(trainMatrix,trainCatergory)

計(jì)算P(i)和P(w[i]|C[1])和P(w[i]|C[0])，這里有兩個(gè)技巧，一個(gè)是開(kāi)始的分子分母沒(méi)有全部初始化為0是為了防止其中一個(gè)的概率為0導(dǎo)致整體為0，另一個(gè)是后面乘用對(duì)數(shù)防止因?yàn)榫葐?wèn)題結(jié)果為0

classifyNB(vec2Classify, p0Vec, p1Vec, pClass1)

根據(jù)貝葉斯公式計(jì)算這個(gè)向量屬于兩個(gè)集合中哪個(gè)的概率高

1.  1 #coding=utf-8 2 from numpy import * 3 def loadDataSet(): 4     postingList=[['my', 'dog', 'has', 'flea', 'PRoblems', 'help', 'please'], 5                  ['maybe', 'not', 'take', 'him', 'to', 'dog', 'park', 'stupid'], 6                  ['my', 'dalmation', 'is', 'so', 'cute', 'I', 'love', 'him'], 7                  ['stop', 'posting', 'stupid', 'worthless', 'garbage'], 8                  ['mr', 'licks', 'ate', 'my', 'steak', 'how', 'to', 'stop', 'him'], 9                  ['quit', 'buying', 'worthless', 'dog', 'food', 'stupid']]10     classVec = [0,1,0,1,0,1]    #1 is abusive, 0 not11     return postingList,classVec12 13 #創(chuàng)建一個(gè)帶有所有單詞的列表14 def createVocabList(dataSet):15     vocabSet = set([])16     for document in dataSet:17         vocabSet = vocabSet | set(document)18     return list(vocabSet)19     20 def setOfWords2Vec(vocabList, inputSet):21     retVocabList = [0] * len(vocabList)22     for word in inputSet:23         if word in vocabList:24             retVocabList[vocabList.index(word)] = 125         else:26             print 'word ',word ,'not in dict'27     return retVocabList28 29 #另一種模型    30 def bagOfWords2VecMN(vocabList, inputSet):31     returnVec = [0]*len(vocabList)32     for word in inputSet:33         if word in vocabList:34             returnVec[vocabList.index(word)] += 135     return returnVec36 37 def trainNB0(trainMatrix,trainCatergory):38     numTrainDoc = len(trainMatrix)39     numWords = len(trainMatrix[0])40     pAbusive = sum(trainCatergory)/float(numTrainDoc)41     #防止多個(gè)概率的成績(jī)當(dāng)中的一個(gè)為042     p0Num = ones(numWords)43     p1Num = ones(numWords)44     p0Denom = 2.045     p1Denom = 2.046     for i in range(numTrainDoc):47         if trainCatergory[i] == 1:48             p1Num +=trainMatrix[i]49             p1Denom += sum(trainMatrix[i])50         else:51             p0Num +=trainMatrix[i]52             p0Denom += sum(trainMatrix[i])53     p1Vect = log(p1Num/p1Denom)#處于精度的考慮，否則很可能到限歸零54     p0Vect = log(p0Num/p0Denom)55     return p0Vect,p1Vect,pAbusive56     57 def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):58     p1 = sum(vec2Classify * p1Vec) + log(pClass1)    #element-wise mult59     p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)60     if p1 > p0:61         return 162     else: 63         return 064         65 def testingNB():66     listOPosts,listClasses = loadDataSet()67     myVocabList = createVocabList(listOPosts)68     trainMat=[]69     for postinDoc in listOPosts:70         trainMat.append(setOfWords2Vec(myVocabList, postinDoc))71     p0V,p1V,pAb = trainNB0(array(trainMat),array(listClasses))72     testEntry = ['love', 'my', 'dalmation']73     thisDoc = array(setOfWords2Vec(myVocabList, testEntry))74     print testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb)75     testEntry = ['stupid', 'garbage']76     thisDoc = array(setOfWords2Vec(myVocabList, testEntry))77     print testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb)78     79     80 def main():81     testingNB()82     83 if __name__ == '__main__':84     main()

    

   機(jī)器學(xué)習(xí)筆記索引