分类算法系列1-----KNN（K近邻）算法思想和python实现

来源：互联网发布：泽宇网络编辑：程序博客网时间：2024/04/29 09:12

1.1K近邻算法的简介

K近邻算法用于测量不同特征值之间的距离进行分类

优点：精度高、对异常值不敏感、无数据输入假设

缺点：计算复杂度高、空间复杂度高

1.2K近邻算计算流程

1.3 K近邻算法例子

下图个K近邻算法的计算步骤流程，由于K=4，Top4有3个为flag1则A的类别为flag1，下图选用最常用的计算距离的公式两点之间距离（欧式距离）

1.5Python代码实现

一下代码参加机器学习实战

from numpy import *import  matplotlib.pyplot  as plt#读数据集def file2matrix(filename):    fr = open(filename)    arror=fr.readlines()    numberOfLines = len(arror)         #get the number of lines in the file    returnMat = zeros((numberOfLines,3))        #prepare matrix to return    classLabelVector = []                       #prepare labels return       index = 0    for line in arror:        line = line.strip()        listFromLine = line.split('\t')        returnMat[index,:] = listFromLine[0:3]        classLabelVector.append(int(listFromLine[-1]))        index += 1    return returnMat,classLabelVector #数据集归一化   def autoNorm(dataSet):    minVals = dataSet.min(0)    maxVals = dataSet.max(0)    ranges = maxVals - minVals    normDataSet = zeros(shape(dataSet))    m = dataSet.shape[0]    normDataSet = dataSet - tile(minVals, (m,1))    normDataSet = normDataSet/tile(ranges, (m,1))   #element wise divide    return normDataSet, ranges, minVals#分类器def classify0(inX, dataSet, labels, k):    dataSetSize = dataSet.shape[0]#数据集记录数    diffMat = tile(inX, (dataSetSize,1)) - dataSet#生成一个和dataSet相同维度的数据集    sqDiffMat = diffMat**2#平方项    sqDistances = sqDiffMat.sum(axis=1)#按行求和    distances = sqDistances**0.5#开方    sortedDistIndicies = distances.argsort()     #对结果进行排序    classCount={}              for i in range(k):        voteIlabel = labels[sortedDistIndicies[i]]        classCount[voteIlabel] = classCount.get(voteIlabel,0) + 1    sortedClassCount = sorted(classCount.iteritems(), key=operator.itemgetter(1), reverse=True)    return sortedClassCount[0][0]#对钱hoRatio*项进行分类，def datingClassTest():    hoRatio = 0.50      #训练集的数据量    datingDataMat,datingLabels = file2matrix('datingTestSet2.txt')       #load data setfrom file    normMat, ranges, minVals = autoNorm(datingDataMat)#数据归一化，normMat归一化的结果，ranges最大值和最小值范围，minVals最小值    m = normMat.shape[0]#行数#    print shape(normMat)    numTestVecs = int(m*hoRatio)#    print numTestVecs    errorCount = 0.0    for i in range(numTestVecs):        classifierResult = classify0(normMat[i,:],normMat[numTestVecs:m,:],datingLabels[numTestVecs:m],5)#        print "the classifier came back with: %d, the real answer is: %d" % (classifierResult, datingLabels[i])        if (classifierResult != datingLabels[i]): errorCount += 1.0    print "the total error rate is: %f" % (errorCount/float(numTestVecs))    print errorCountdef classifyPerson():    resultList=['not at all','in some doses','in large doses']    ffmiles=float(raw_input('frequent filer miles earned'))    precents=float(raw_input('precents filer miles earned'))    iceCream=float(raw_input('liters of ice'))    datingDataMat,datingLabels = file2matrix('datingTestSet2.txt')     normMat, ranges, minVals = autoNorm(datingDataMat)    inArray=[ffmiles,precents,iceCream]    classifierResult = classify0((inArray-minVals)*1.0/ranges,normMat,datingLabels,3)      print resultList[classifierResult-1]    print classifierResultif __name__=='__main__':    datingClassTest()    classifyPerson()

下图为数据的分布趋势

下面为用户手动输入数据观测数据所属类别

至此K近邻算法的实现基本完成

0 0