机器学习之KNN算法

1. 写在前面的话

去年学完python之后，就再也没用过，一直在做软件相关的项目，已经快烦死了 。最近稍微可以做点自己的事，准备开始自己的机器学习之路（其实就想做做和数据相关的工作，什么数据分析之类的，然后想到了机器学习，毕竟本科是数学专业），好了废话就这么多了，代码敲起。

2.1 了解KNN算法

简单的说，就是求每个测试样本与训练集之间的欧氏距离，然后根据距离的远近来分类的算法，其伪代码如下：

• 计算已知类别数据集中的点与当前点的欧氏距离
• 按照距离递增进行排序
• 选取与当前点的距离最小的k个点
• 确定前k个点所在类别出现的频率
• 返回前k个点出现频率最干的类别作为当前点的预测分类

2.2 具体代码

我们以约会网站的 数据进行测试，看看使用KNN算法进行分类的效果是否理想。

import numpy as np
import operator

# 产生测试用列
def createDataSet():
    group = np.array([[1.0, 1.1], [1.0, 1.0], [0, 0], [0, 0.1]])
    labels = ['A', 'A', 'B', 'B']
    return group, labels

# 分类函数（KNN算法实现）
def classify(inX, dataSet, labels, k):
    dataSetSize = dataSet.shape[0]
    # np.tile() 复制函数 沿y轴复制 dataSetSize 次
    diffMat = np.tile(inX, (dataSetSize, 1)) - dataSet
    sqDiffMat = diffMat ** 2
    sqDistance = sqDiffMat.sum(axis=1)
    distances = sqDistance ** 0.5
    sortedDistIndicies = np.argsort(distances)
    
    classCount = {}
    for i in range(k):
        voteIlable = labels[sortedDistIndicies[i]]
        classCount[voteIlable] = classCount.get(voteIlable, 0) + 1
    sortedClassCount = sorted(classCount.items(), key=operator.itemgetter(1), reverse=True)
    return sortedClassCount[0][0]

# 获取数据
def file2matrix(filename):
    fr = open(filename)
    # 获取所有行
    arrayOLines = fr.readlines()
    numberOfLines = len(arrayOLines)
    # 产生一个 numberOfLines 行3列的矩阵
    returnMat = np.zeros((numberOfLines, 3))
    classLabelVector = []
    for index in range(numberOfLines):
        listFromLine = arrayOLines[index].strip().split('\t')
        returnMat[index, :] = listFromLine[0:3]
        classLabelVector.append(int(listFromLine[-1]))
        
    return returnMat, classLabelVector

# 数据归一化
def autoNorm(dataSet):
    # 每一列的最小值，列表形式
    minVals = dataSet.min(0)
    # 每一列的最大值，列表形式
    maxVals = dataSet.max(0)
    # 区间值，列表形式
    ranges = maxVals - minVals
    normDataSet = np.zeros(np.shape(dataSet))
    m = dataSet.shape[0]
    normDataSet = dataSet- np.tile(minVals, (m, 1))
    normDataSet = normDataSet / np.tile(ranges, (m, 1))
    
    return normDataSet, ranges, minVals

# 测试函数
def datingClassTest():
    hoRatio = 0.05
    datingDataMat, datingLabels = file2matrix('datingTestSet2.txt')
    normMat, ranges, minVals = autoNorm(datingDataMat)
    m = normMat.shape[0]
    numTestVecs = int(m * hoRatio)
    errorCount = 0.0
    for i in range(numTestVecs):
        classifierResult = classify(normMat[i, :], normMat[numTestVecs:m, :], datingLabels[numTestVecs:m], 3)
        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)))
    
datingClassTest()

我们使用总数据的5%来进行测试，用95%的数据来进行训练，得到的错误在2%，已经非常理想，效果见下。

runfile('/Volumes/文档/ML/KNN.py', wdir='/Volumes/文档/ML')
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 3, the real answer is: 2
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 2, the real answer is: 2
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 3, the real answer is: 3
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 1, the real answer is: 1
The classifier came back with: 2, the real answer is: 2
The total error rate is: 0.020000