【Machine Learning】 朴素贝叶斯
1. 朴素贝叶斯: 条件概率在机器学习算法的应用。理解这个算法需要一点推导。不会编辑公式。。
核心就是 在已知训练集的前提条件下,算出每个特征的概率为该分类的概率, 然后套贝叶斯公式计算 预测集的所有分类概率,预测类型为概率最大的类型
from numpy import * def loadDataSet(): """ Returns: postingList: list, 用于测试的静态数据 classVec: list, 标签,与 postingList 对应, 1 代表侮辱性文字, 0 代表正常言论 """ 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 代表正常言论 return postingList, classVec def createVocabList(dataSet): """ 对数据进行去重 Args: dataSet: list, 原始数据 Returns: list 去重后的一维list """ vocabSet = set([]) for document in dataSet: vocabSet = vocabSet | set(document) return list(vocabSet) def setOfWords2Vec(vocabList, inputSet): """ 对数据使用情况进行标记 Args: vocabList: list 参考数据 inputSet: 测试数据 Returns: list 对应 vocabList, 1 代表在 inputSet 中存在, 0 代表不存在 """ returnVec = [0] * len(vocabList) for word in inputSet: if word in inputSet: returnVec[vocabList.index(word)] = 1 else: print("the word: %s is not my Vocabulary!" % word) return returnVec def trainNB0(trainMatrix, trainCategory): """ Args: trainMatrix: 测试数据 trainCategory: 数据标签 Returns: p0Vect: list[list] 在已知正常文档的概率是 0.4的前提下, 每个单词的为正常单词的概率) p1Vect: list[list] (在已知侮辱性文档的概率是 0.6的前提下, 每个单词的为侮辱性单词的概率) pAbusive: float 条件概率中的条件 以 createDataSet 方法中的数据为例, 侮辱性文档的概率是 0.6, 正常文档的概率是0.4 """ numTrainDocs = len(trainMatrix) numWords = len(trainMatrix[0]) pAbusive = sum(trainCategory) / float(numTrainDocs) p0Num = zeros(numWords) p1Num = zeros(numWords) p0Denom = 0.0 p1Denom = 0.0 for i in range(numTrainDocs): if trainCategory[i] == 1: p1Num += trainMatrix[i] p1Denom += sum(trainMatrix[i]) else: p0Num += trainMatrix[i] p0Denom += sum(trainMatrix[i]) p1Vect = p1Num / p1Denom p0Vect = p0Num / p0Denom return p0Vect, p1Vect, pAbusive def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1): """ Args: vec2Classify: list[int] 词汇表使用标记 p0Vec: list[list] 单个词汇为正常词汇的概率 p1Vec: list[list] 单个词汇为侮辱性词汇的概率 pClass1: float 文档为侮辱性文档的概率 Returns: 1: 侮辱性文档 2: 正常文档 """ p1 = sum(vec2Classify * p1Vec) + log(pClass1) # 使用log避免 0乘以任何数为0 的尴尬 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(trainMat, 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)) if __name__ == '__main__': testingNB()
欢迎访问我的个人博客站点:
https://yeyeck.com