机器学习实战之KNN

KNN也称K-近邻算法，简单来说，KNN采用测量不同特征值之间的距离的方法进行分类。

优点：精度高，对异常值不敏感，无数据输入假定。

确定：时间复杂度、空间复杂度较高

适用数据范围：数值型和标称型

工作原理简介：存在一个样本数据集合，并且每个样本数据集中都存在标签。输入没有标签的数据集之后，将新数据集的每个特征与样本集中数据对应的特征进行比较，然后算法提取样本集中特征最相似数据（最近邻）的分类标签。一般来说，我们只选择样本中钱K个最相似的数据，这就是K-近邻中K的出处，通常K是不大于20的整数。最后，选择K个最相似数据中出现次数最多的分类，作为新数据集的分类标签。

示例1：使用KNN改进约会网站的配对效果

labels:不喜欢的人

　　   魅力一般的人

　　   极具魅力的人

Feature：每年获得的飞行常客里程数

　　　　玩视频游戏所消耗的时间百分比

　　　　每周消耗的冰淇淋公升数

示例代码如下：

import numpy as np
import operator
import matplotlib
import matplotlib.pyplot as plt

#KNN 分类器
def classify0(inX, dataSet, labels, k):
    dataSetSize = dataSet.shape[0]
    diffMat = np.tile(inX, (dataSetSize, 1)) - dataSet #复制inX为（dataSetSize, 1）
    sqDiffMat = diffMat ** 2
    sqDistances = sqDiffMat.sum(axis = 1)
    distances = sqDistances ** 0.5
    sortedDistIndicies = distances.argsort() #按照value排序，并且返回索引
    classCount = {}
    for i in range(k):
        voteIlabel = labels[sortedDistIndicies[i]]
        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1

    sortedClassCount = sorted(classCount.items(), key = lambda item: item[1] ,reverse = True)

    return sortedClassCount[0][0]

#将文件中数据转化为矩阵数据
def file2matrix(fileName):
    fr = open(fileName)
    arrayAllLines = fr.readlines()
    numberOfLines = len(arrayAllLines)
    returnMat = np.zeros((numberOfLines, 3))
    classLabelVector = []
    index = 0

    for line in arrayAllLines:
        line.strip() #默认删除此行开头和结尾的空格和换行
        listFromLine = line.split('\t')
        returnMat[index,:] = listFromLine[0:3]
        classLabelVector.append(int(listFromLine[-1]))
        index += 1

    return returnMat, classLabelVector

#归一化
def autoNorm(dataSet):
    minVals = dataSet.min(0)
    maxVals = dataSet.max(0)
    ranges = maxVals - minVals
    row = dataSet.shape[0]
    normDataSet = dataSet - np.tile(minVals, (row, 1))
    normDataSet = normDataSet / np.tile(ranges, (row, 1))
    return normDataSet, ranges, minVals

#随机选取10%的数据进行测试
def datingClassTest():
    hoRatio = 0.10
    pathName = "./datingTestSet2.txt"
    datingDataMat, datingLabels = file2matrix(pathName)
    normMat, ranges, minVals = autoNorm(datingDataMat)
    row = normMat.shape[0]
    numTestVecs = int(hoRatio * row)
    errorCount = 0.0
    for i in range(numTestVecs):
        classifierResult = classify0(normMat[i,:], normMat[numTestVecs:row,:], datingLabels[numTestVecs:row], 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 tate is : %f" % (errorCount / float(numTestVecs)))

datingClassTest()

示例2：手写数字识别系统

import numpy as np
import operator
import matplotlib
import matplotlib.pyplot as plt
from os import listdir

#KNN 分类器
def classify0(inX, dataSet, labels, k):
    dataSetSize = dataSet.shape[0]
    diffMat = np.tile(inX, (dataSetSize, 1)) - dataSet #复制inX为（dataSetSize, 1）
    sqDiffMat = diffMat ** 2
    sqDistances = sqDiffMat.sum(axis = 1)
    distances = sqDistances ** 0.5
    sortedDistIndicies = distances.argsort() #按照value排序，并且返回索引
    classCount = {}
    for i in range(k):
        voteIlabel = labels[sortedDistIndicies[i]]
        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1

    sortedClassCount = sorted(classCount.items(), key = lambda item: item[1] ,reverse = True)

    return sortedClassCount[0][0]

#将文本文件转化为向量
def img2vector(fileName):
    returnVect = np.zeros((1,32 * 32))
    fr = open(fileName)
    for i in range(32):
        lineStr = fr.readline()
        for j in range(32):
            returnVect[0,i * 32 + j] = int(lineStr[j])

    return returnVect

#手写数字测试错误率
def handwritingClassTest():
    hwLabels = []
    trainingFileList = listdir('trainingDigits')
    m = len(trainingFileList)
    trainingMat = np.zeros((m, 32 * 32))
    for i in range(m):
        fileNameStr = trainingFileList[i]
        fileStr = fileNameStr.split('.')[0]
        classNum = int(fileStr.split('_')[0])
        hwLabels.append(classNum)
        trainingMat[i,:] = img2vector('trainingDigits/%s' % fileNameStr)

    testFileList = listdir('testDigits')
    m = len(testFileList)
    errorCount = 0.0
    for i in range(m):
        fileNameStr = testFileList[i]
        fileStr = fileNameStr.split('.')[0]
        realClassNum = int(fileStr.split('_')[0])
        testVect = img2vector('testDigits/%s' % fileNameStr)
        testClassNum = classify0(testVect, trainingMat, hwLabels, 3)
        print("The classifier came back with: %d, the real answer is: %d" % (testClassNum, realClassNum))
        if(testClassNum != realClassNum):
            errorCount += 1.0

    print("The total error rate is: %f" % (errorCount / float(m)))

handwritingClassTest()

数据集下载以及完整jupyter notebook 代码下载：

https://github.com/qwqwqw110/machineLearningInactionCode/tree/master/KNN