朴素贝叶斯算法的python实现
算法优缺点
利益:在数据较少的环境下依然有效,可以处理惩罚多种别问题
缺点:对输入数据的筹备方法敏感
合用数据范例:标称型数据
算法思想:
朴素贝叶斯
好比我们想判定一个邮件是不是垃圾邮件,那么我们知道的是这个邮件中的词的漫衍,那么我们还要知道:垃圾邮件中某些词的呈现是几多,就可以操作贝叶斯定理获得。
朴素贝叶斯分类器中的一个假设是:每个特征同等重要
贝叶斯分类是一类分类算法的总称,这类算法均以贝叶斯定理为基本,故统称为贝叶斯分类。
函数
loadDataSet()
建设数据集,这里的数据集是已经拆分好的单词构成的句子,暗示的是某论坛的用户评论,标签1暗示这个是骂人的
createVocabList(dataSet)
找出这些句子中总共有几多单词,以确定我们词向量的巨细
setOfWords2Vec(vocabList, inputSet)
将句子按照个中的单词转成向量,这里用的是伯努利模子,即只思量这个单词是否存在
bagOfWords2VecMN(vocabList, inputSet)
这个是将句子转成向量的另一种模子,多项式模子,思量某个词的呈现次数
trainNB0(trainMatrix,trainCatergory)
计较P(i)和P(w[i]|C[1])和P(w[i]|C[0]),这里有两个能力,一个是开始的分子分母没有全部初始化为0是为了防备个中一个的概率为0导致整体为0,另一个是后头乘用对数防备因为精度问题功效为0
classifyNB(vec2Classify, p0Vec, p1Vec, pClass1)
按照贝叶斯公式计较这个向量属于两个荟萃中哪个的概率高
#coding=utf-8 from numpy 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 is abusive, 0 not return postingList,classVec #建设一个带有所有单词的列表 def createVocabList(dataSet): vocabSet = set([]) for document in dataSet: vocabSet = vocabSet | set(document) return list(vocabSet) def setOfWords2Vec(vocabList, inputSet): retVocabList = [0] * len(vocabList) for word in inputSet: if word in vocabList: retVocabList[vocabList.index(word)] = 1 else: print 'word ',word ,'not in dict' return retVocabList #另一种模子 def bagOfWords2VecMN(vocabList, inputSet): returnVec = [0]*len(vocabList) for word in inputSet: if word in vocabList: returnVec[vocabList.index(word)] += 1 return returnVec def trainNB0(trainMatrix,trainCatergory): numTrainDoc = len(trainMatrix) numWords = len(trainMatrix[0]) pAbusive = sum(trainCatergory)/float(numTrainDoc) #防备多个概率的后果傍边的一个为0 p0Num = ones(numWords) p1Num = ones(numWords) p0Denom = 2.0 p1Denom = 2.0 for i in range(numTrainDoc): if trainCatergory[i] == 1: p1Num +=trainMatrix[i] p1Denom += sum(trainMatrix[i]) else: p0Num +=trainMatrix[i] p0Denom += sum(trainMatrix[i]) p1Vect = log(p1Num/p1Denom)#处于精度的思量,不然很大概到限归零 p0Vect = log(p0Num/p0Denom) return p0Vect,p1Vect,pAbusive def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1): p1 = sum(vec2Classify * p1Vec) + log(pClass1) #element-wise mult p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1) if p1 > p0: return 1 else: return 0 def testingNB(): listOPosts,listClasses = loadDataSet() myVocabList = createVocabList(listOPosts) trainMat=[] for postinDoc in listOPosts: trainMat.append(setOfWords2Vec(myVocabList, postinDoc)) p0V,p1V,pAb = trainNB0(array(trainMat),array(listClasses)) testEntry = ['love', 'my', 'dalmation'] thisDoc = array(setOfWords2Vec(myVocabList, testEntry)) print testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb) testEntry = ['stupid', 'garbage'] thisDoc = array(setOfWords2Vec(myVocabList, testEntry)) print testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb) def main(): testingNB() if __name__ == '__main__': main()