朴素贝叶斯分类器 注释
试编程实现拉普拉斯修正的朴素贝叶斯分类器，并以西瓜数据集3.0为训练集，对P.151“测1”进行判别。
代码全是《机器学习》上的，只是将其整合到了一起，能够运行手写体识别。
内容大部分进行了注释，可能有些注释不够精准或者不容理解，见谅！
代码
from numpy import *
def loadDataSet(): #创建实验样本
postingList=[['青绿','蜷缩','浊响','清晰','凹陷','硬滑',0.697,0.460],
['乌黑','蜷缩','沉闷','清晰','凹陷','硬滑',0.774,0.376],
['乌黑','蜷缩','浊响','清晰','凹陷','硬滑',0.634,0.264],
['青绿','蜷缩','沉闷','清晰','凹陷','硬滑',0.608,0.318],
['浅白','蜷缩','浊响','清晰','凹陷','硬滑',0.556,0.215],
['青绿','稍蜷','浊响','清晰','稍凹','软粘',0.403,0.237],
['乌黑','稍蜷','浊响','稍糊','稍凹','软粘',0.481,0.149],
['乌黑','稍蜷','浊响','清晰','稍凹','硬滑',0.437,0.211],
['乌黑','稍蜷','沉闷','稍糊','稍凹','硬滑',0.666,0.091],
['青绿','硬挺','清脆','清晰','平坦','软粘',0.243,0.267],
['浅白','硬挺','清脆','模糊','平坦','硬滑',0.245,0.057],
['浅白','蜷缩','浊响','模糊','平坦','软粘',0.343,0.099],
['青绿','稍蜷','浊响','稍糊','凹陷','硬滑',0.639,0.161],
['浅白','稍蜷','沉闷','稍糊','凹陷','硬滑',0.657,0.198],
['乌黑','稍蜷','浊响','清晰','稍凹','软粘',0.360,0.370],
['浅白','蜷缩','浊响','模糊','平坦','硬滑',0.593,0.042],
['青绿','蜷缩','沉闷','稍糊','稍凹','硬滑',0.719,0.103]]
classVec = [1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0,0] #1是好瓜，0不是好瓜
return postingList,classVec
def createVocabList(dataSet): #传入词条切分后的列表
vocabSet = set([]) #用set（）创建空的集合
for document in dataSet:
vocabSet = vocabSet | set(document) #合并两个集合
return list(vocabSet)
def setOfWords2Vec(vocabList, inputSet): #传入一个集合和被检测列表
returnVec = [0]*len(vocabList) #初始化列表全为0
for word in inputSet: #遍历被检测列表
if word in vocabList: #检测被检测列表中是否有给定样本中的词汇
returnVec[vocabList.index(word)] = 1 #将该词出现的第一个位置的特征值标记为1
else: print( "the word: %s is not in my Vocabulary!" % word)
return returnVec
def trainNB0(trainMatrix,trainCategory): #拉普斯修改后的朴素贝叶斯算法
numTrainDocs = len(trainMatrix)
numWords = len(trainMatrix[0])
pAbusive = sum(trainCategory)/float(numTrainDocs) #计算坏瓜所占比例
p0Num = ones(numWords); p1Num = ones(numWords)
p0Denom = 2.0; p1Denom = 2.0
for i in range(numTrainDocs): #遍历每个文本
if trainCategory[i] == 1:
p1Num += trainMatrix[i]
p1Denom += sum(trainMatrix[i])
else:
p0Num += trainMatrix[i]
p0Denom += sum(trainMatrix[i])
p1Vect = log(p1Num/p1Denom) #计算每个属性被判为坏瓜属性的概率
p0Vect = log(p0Num/p0Denom) #基本同上，避免结果下溢
return p0Vect,p1Vect,pAbusive
def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):
p1 = sum(vec2Classify * p1Vec) + log(pClass1) #属性在出现0/1情况*概率 再求和 + 类别对数概率
p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)
if p1 > p0:
return 1
else:
return 0
def bagOfWords2VecMN(vocabList, inputSet):
returnVec = [0]*len(vocabList)
for word in inputSet:
if word in vocabList:
returnVec[vocabList.index(word)] += 1 #每个单词再集合中的出现次数
return returnVec
def testingNB(canshu):
listOPosts,listClasses = loadDataSet() #取样
myVocabList = createVocabList(listOPosts) #变为无重复集合
trainMat=[]
for postinDoc in listOPosts: #标记每个属性是否再大集合中出现
trainMat.append(setOfWords2Vec(myVocabList, postinDoc))
p0V,p1V,pAb = trainNB0(array(trainMat),array(listClasses))
testEntry = canshu
thisDoc = array(setOfWords2Vec(myVocabList, testEntry)) #标记testEntry属性再集合中的出现
print (testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb))
if __name__=='__main__':
canshu=list(input("请输入:").split()) #将结果输入
testingNB(canshu)
运行结果
