KNN算法，k 鄰近(python)

1.原理
利用歐式距離計算各個特徵的相似度，歐式距離越小，相似度越大。

2.代碼：

from numpy import *
import operator
from os import listdir

def kNNClassify(inX, dataSet, labels, k):
    #sample size
    dataSize = dataSet.shape[0]
    #get the difference between inX and sample
    diffMat = tile(inX, (dataSize,1)) - dataSet
    diffMat = diffMat**2
    #get sum of each row so set axis = 1
    sumMat = diffMat.sum(axis = 1)
    sqdiffMat = sumMat**0.5
    #sort sqdiffMat and get the index result
    diffSortIndices = sqdiffMat.argsort()
    #res dict
    resLabels = {}
    maxTimes = 0
    for i range(k):
        tempLabel = labels[diffSortIndices[i]]
        resLabels[tempLabel] = resLabels.get(tempLabel, 0) + 1

    for key in resLabels.keys():
        if(resLabels.get(key) > maxTimes):
            maxTimes = resLabels.get(key)
            res = key
    return res