KNN算法实现

KNN算法实现

一、实验名称：
KNN算法实现
二、实验目的：
掌握KNN算法基本原理
三、实验要求
1、使用python实现KNN；
2、测试不同的k值下KNN算法分类效果；

四、实验平台
计算机、Python、Anaconda
五、实验数据
processed.cleveland.data
UCI公开数据集-heart disease，属性信息如下:

1. #3 (age)
2. #4 (sex)
3. #9 (cp)
4. #10 (trestbps)
5. #12 (chol)
6. #16 (fbs)
7. #19 (restecg)
8. #32 (thalach)
9. #38 (exang)
10. #40 (oldpeak)
11. #41 (slope)
12. #44 (ca)
13. #51 (thal)
14. #58 (num) (the predicted attribute)

数据集参考网址：https://archive.ics.uci.edu/ml/datasets/Heart+Disease

1.在数据集页面，下载本实验所需数据集 。

2.实验代码

import numpy as np

f=open(r'E:\360MoveData\Users\DELL\Desktop\processed.cleveland.data')
age=[]
sex=[]
cp=[]
tresbps=[]
chol=[]
fbs=[]
restecg=[]
thalach=[]
exang=[]
oldpeak=[]
slope=[]
ca=[]
thal=[]
num=[]
for i,d in enumerate(f):
    d=d.strip()
    if not d:
        continue
    d=list(map(float,d.split(',')))
    age.append(d[0])
    sex.append(d[1])
    cp.append(d[2])
    tresbps.append(d[3])
    chol.append(d[4])
    fbs.append(d[5])
    restecg.append(d[6])
    thalach.append(d[7])
    exang.append(d[8])
    oldpeak.append(d[9])
    slope.append(d[10])
    ca.append(d[11])
    thal.append(d[12])
    num.append(d[13])
group=np.empty(shape=[0,13],dtype=float)
labels=np.empty(shape=[0,1],dtype=float)
for i in range(0,303):
    group=np.append(group,[[age[i],sex[i],cp[i],tresbps[i],chol[i],fbs[i],restecg[i],thalach[i],exang[i],oldpeak[i],slope[i],ca[i],thal[i]]],axis=0)
    labels=np.append(labels,[[num[i]]],axis=0)
        
def kNN_Classify(inX, dataSet, labels, k):
    dataSetSize = dataSet.shape[0]
    diffMat = np.tile(inX, (dataSetSize, 1)) - dataSet
    #关于tile函数的用法
    #>>> b=[1,3,5]
    #>>> tile(b,[2,3])
    #array([[1, 3, 5, 1, 3, 5, 1, 3, 5],
    #       [1, 3, 5, 1, 3, 5, 1, 3, 5]])
    sqDiffMat = diffMat ** 2
    sqDistances = sum(sqDiffMat)
    distances = sqDistances ** 0.5            #            算距离
    sortedDistIndicies =np.argsort(distances)
    #关于argsort函数的用法
    #argsort函数返回的是数组值从小到大的索引值
    #>>> x = np.array([3, 1, 2])
    #>>> np.argsort(x)
    #array([1, 2, 0])
    classCount = {}                                          #定义一个字典
#   选择k个最近邻
    for i in range(k):
        voteLabel = labels[sortedDistIndicies[i]]
                                                            # 计算k个最近邻中各类别出现的次数
        classCount[voteLabel[i]] = sortedDistIndicies[i] + 1
 
    #返回出现次数最多的类别标签i
    maxCount = 0
    for key, value in classCount.items():
        if value > maxCount:
            maxCount = value
            maxIndex = key
    return maxIndex
 
dataSet=group
test=np.array([61.0,1.0,1.0,134.0,234.0,0.0,0.0,145.0,0.0,2.6,2.0,2.0,3.0])
print("请输入K值")
k=int(input())
outputLabel = kNN_Classify(test, dataSet, labels,k)
print("输入的样本", test, "心血管病的类别为 ", outputLabel)


 

3.实验结果

注：一些读代码的理解过程

test=np.array([61.0,1.0,1.0,134.0,234.0,0.0,0.0,145.0,0.0,2.6,2.0,2.0,3.0])
dataSetSize = 303
diffMat = np.tile(test, (dataSetSize, 1)) - group
print(diffMat)
sqDiffMat = diffMat ** 2
print(sqDiffMat)
sqDistances = sum(sqDiffMat)
print(sqDistances)
distances = sqDistances ** 0.5      
print(distances)    

sortedDistIndicies =np.argsort(distances)
print(sortedDistIndicies)#从小到大距离下标
k=1
classCount = {}
for i in range(k):
        voteLabel = labels[sortedDistIndicies[i]] #距离最小的结果
        print(voteLabel)#结果值
        a = sortedDistIndicies[i] + 1
        print(a)
        classCount[voteLabel[i]] = sortedDistIndicies[i] + 1
        print(classCount)
        
        maxCount = 0
for key, value in classCount.items():
            if value > maxCount:
               maxCount = value
               maxIndex = key
return maxIndex

参考：https://blog.csdn.net/wxwmb11/article/details/106447279，做了一些修改，感谢。