快捷導(dǎo)航

python實(shí)現(xiàn)KNN近鄰算法

更新時(shí)間：2020年12月30日 10:21:47 作者：呱唧_T_呱唧

這篇文章主要介紹了python實(shí)現(xiàn)KNN近鄰算法的方法，幫助大家更好的利用python進(jìn)行機(jī)器學(xué)習(xí)，感興趣的朋友可以了解下

示例：《電影類型分類》

獲取數(shù)據(jù)來源

電影名稱	打斗次數(shù)	接吻次數(shù)	電影類型
California Man	3	104	Romance
He's Not Really into Dudes	8	95	Romance
Beautiful Woman	1	81	Romance
Kevin Longblade	111	15	Action
Roob Slayer 3000	99	2	Action
Amped II	88	10	Action
Unknown	18	90	unknown

數(shù)據(jù)顯示：肉眼判斷電影類型unknown是什么

from matplotlib import pyplot as plt

# 用來正常顯示中文標(biāo)簽
plt.rcParams["font.sans-serif"] = ["SimHei"]
# 電影名稱
names = ["California Man", "He's Not Really into Dudes", "Beautiful Woman",
   "Kevin Longblade", "Robo Slayer 3000", "Amped II", "Unknown"]
# 類型標(biāo)簽
labels = ["Romance", "Romance", "Romance", "Action", "Action", "Action", "Unknown"]
colors = ["darkblue", "red", "green"]
colorDict = {label: color for (label, color) in zip(set(labels), colors)}
print(colorDict)
# 打斗次數(shù)，接吻次數(shù)
X = [3, 8, 1, 111, 99, 88, 18]
Y = [104, 95, 81, 15, 2, 10, 88]

plt.title("通過打斗次數(shù)和接吻次數(shù)判斷電影類型", fontsize=18)
plt.xlabel("電影中打斗鏡頭出現(xiàn)的次數(shù)", fontsize=16)
plt.ylabel("電影中接吻鏡頭出現(xiàn)的次數(shù)", fontsize=16)

# 繪制數(shù)據(jù)
for i in range(len(X)):
 # 散點(diǎn)圖繪制
 plt.scatter(X[i], Y[i], color=colorDict[labels[i]])

# 每個(gè)點(diǎn)增加描述信息
for i in range(0, 7):
 plt.text(X[i]+2, Y[i]-1, names[i], fontsize=14)

plt.show()

問題分析：根據(jù)已知信息分析電影類型unknown是什么

核心思想：

未標(biāo)記樣本的類別由距離其最近的K個(gè)鄰居的類別決定

距離度量：

一般距離計(jì)算使用歐式距離（用勾股定理計(jì)算距離），也可以采用曼哈頓距離（水平上和垂直上的距離之和）、余弦值和相似度（這是距離的另一種表達(dá)方式）。相比于上述距離，馬氏距離更為精確，因?yàn)樗芸紤]很多因素，比如單位，由于在求協(xié)方差矩陣逆矩陣的過程中，可能不存在，而且若碰見3維及3維以上，求解過程中極其復(fù)雜，故可不使用馬氏距離

知識擴(kuò)展

馬氏距離概念：表示數(shù)據(jù)的協(xié)方差距離
方差：數(shù)據(jù)集中各個(gè)點(diǎn)到均值點(diǎn)的距離的平方的平均值
標(biāo)準(zhǔn)差：方差的開方
協(xié)方差cov(x, y)：E表示均值，D表示方差，x，y表示不同的數(shù)據(jù)集，xy表示數(shù)據(jù)集元素對應(yīng)乘積組成數(shù)據(jù)集

cov(x, y) = E(xy) - E(x)*E(y)

cov(x, x) = D(x)

cov(x1+x2, y) = cov(x1, y) + cov(x2, y)

cov(ax, by) = abcov(x, y)

協(xié)方差矩陣：根據(jù)維度組成的矩陣，假設(shè)有三個(gè)維度，a，b，c

∑ij = [cov(a, a) cov(a, b) cov(a, c) cov(b, a) cov(b,b) cov(b, c) cov(c, a) cov(c, b) cov(c, c)]

算法實(shí)現(xiàn)：歐氏距離

編碼實(shí)現(xiàn)

# 自定義實(shí)現(xiàn) mytest1.py
import numpy as np

# 創(chuàng)建數(shù)據(jù)集
def createDataSet():
 features = np.array([[3, 104], [8, 95], [1, 81], [111, 15],
       [99, 2], [88, 10]])
 labels = ["Romance", "Romance", "Romance", "Action", "Action", "Action"]
 return features, labels

def knnClassify(testFeature, trainingSet, labels, k):
 """
 KNN算法實(shí)現(xiàn)，采用歐式距離
 :param testFeature: 測試數(shù)據(jù)集，ndarray類型，一維數(shù)組
 :param trainingSet: 訓(xùn)練數(shù)據(jù)集，ndarray類型，二維數(shù)組
 :param labels: 訓(xùn)練集對應(yīng)標(biāo)簽，ndarray類型，一維數(shù)組
 :param k: k值，int類型
 :return: 預(yù)測結(jié)果，類型與標(biāo)簽中元素一致
 """
 dataSetsize = trainingSet.shape[0]
 """
 構(gòu)建一個(gè)由dataSet[i] - testFeature的新的數(shù)據(jù)集diffMat
 diffMat中的每個(gè)元素都是dataSet中每個(gè)特征與testFeature的差值（歐式距離中差）
 """
 testFeatureArray = np.tile(testFeature, (dataSetsize, 1))
 diffMat = testFeatureArray - trainingSet
 # 對每個(gè)差值求平方
 sqDiffMat = diffMat ** 2
 # 計(jì)算dataSet中每個(gè)屬性與testFeature的差的平方的和
 sqDistances = sqDiffMat.sum(axis=1)
 # 計(jì)算每個(gè)feature與testFeature之間的歐式距離
 distances = sqDistances ** 0.5

 """
 排序，按照從小到大的順序記錄distances中各個(gè)數(shù)據(jù)的位置
 如distance = [5, 9, 0, 2]
 則sortedStance = [2, 3, 0, 1]
 """
 sortedDistances = distances.argsort()

 # 選擇距離最小的k個(gè)點(diǎn)
 classCount = {}
 for i in range(k):
  voteiLabel = labels[list(sortedDistances).index(i)]
  classCount[voteiLabel] = classCount.get(voteiLabel, 0) + 1
 # 對k個(gè)結(jié)果進(jìn)行統(tǒng)計(jì)、排序，選取最終結(jié)果，將字典按照value值從大到小排序
 sortedclassCount = sorted(classCount.items(), key=lambda x: x[1], reverse=True)
 return sortedclassCount[0][0]

testFeature = np.array([100, 200])
features, labels = createDataSet()
res = knnClassify(testFeature, features, labels, 3)
print(res)
# 使用python包實(shí)現(xiàn) mytest2.py
from sklearn.neighbors import KNeighborsClassifier
from .mytest1 import createDataSet

features, labels = createDataSet()
k = 5
clf = KNeighborsClassifier(k_neighbors=k)
clf.fit(features, labels)

# 樣本值
my_sample = [[18, 90]]
res = clf.predict(my_sample)
print(res)

示例：《交友網(wǎng)站匹配效果預(yù)測》

數(shù)據(jù)來源：略

數(shù)據(jù)顯示

import pandas as pd
import numpy as np
from matplotlib import pyplot as plt
from mpl_toolkits.mplot3d import Axes3D

# 數(shù)據(jù)加載
def loadDatingData(file):
 datingData = pd.read_table(file, header=None)
 datingData.columns = ["FlightDistance", "PlaytimePreweek", "IcecreamCostPreweek", "label"]
 datingTrainData = np.array(datingData[["FlightDistance", "PlaytimePreweek", "IcecreamCostPreweek"]])
 datingTrainLabel = np.array(datingData["label"])
 return datingData, datingTrainData, datingTrainLabel

# 3D圖顯示數(shù)據(jù)
def dataView3D(datingTrainData, datingTrainLabel):
 plt.figure(1, figsize=(8, 3))
 plt.subplot(111, projection="3d")
 plt.scatter(np.array([datingTrainData[x][0]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "smallDoses"]),
    np.array([datingTrainData[x][1]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "smallDoses"]),
    np.array([datingTrainData[x][2]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "smallDoses"]), c="red")
 plt.scatter(np.array([datingTrainData[x][0]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "didntLike"]),
    np.array([datingTrainData[x][1]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "didntLike"]),
    np.array([datingTrainData[x][2]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "didntLike"]), c="green")
 plt.scatter(np.array([datingTrainData[x][0]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "largeDoses"]),
    np.array([datingTrainData[x][1]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "largeDoses"]),
    np.array([datingTrainData[x][2]
       for x in range(len(datingTrainLabel))
       if datingTrainLabel[x] == "largeDoses"]), c="blue")
 plt.xlabel("飛行里程數(shù)", fontsize=16)
 plt.ylabel("視頻游戲耗時(shí)百分比", fontsize=16)
 plt.clabel("冰淇凌消耗", fontsize=16)
 plt.show()
 
datingData, datingTrainData, datingTrainLabel = loadDatingData(FILEPATH1)
datingView3D(datingTrainData, datingTrainLabel)

問題分析：抽取數(shù)據(jù)集的前10%在數(shù)據(jù)集的后90%進(jìn)行測試

編碼實(shí)現(xiàn)

# 自定義方法實(shí)現(xiàn)
import pandas as pd
import numpy as np

# 數(shù)據(jù)加載
def loadDatingData(file):
 datingData = pd.read_table(file, header=None)
 datingData.columns = ["FlightDistance", "PlaytimePreweek", "IcecreamCostPreweek", "label"]
 datingTrainData = np.array(datingData[["FlightDistance", "PlaytimePreweek", "IcecreamCostPreweek"]])
 datingTrainLabel = np.array(datingData["label"])
 return datingData, datingTrainData, datingTrainLabel

# 數(shù)據(jù)歸一化
def autoNorm(datingTrainData):
 # 獲取數(shù)據(jù)集每一列的最值
 minValues, maxValues = datingTrainData.min(0), datingTrainData.max(0)
 diffValues = maxValues - minValues
 
 # 定義形狀和datingTrainData相似的最小值矩陣和差值矩陣
 m = datingTrainData.shape(0)
 minValuesData = np.tile(minValues, (m, 1))
 diffValuesData = np.tile(diffValues, (m, 1))
 normValuesData = (datingTrainData-minValuesData)/diffValuesData
 return normValuesData

# 核心算法實(shí)現(xiàn)
def KNNClassifier(testData, trainData, trainLabel, k):
 m = trainData.shape(0)
 testDataArray = np.tile(testData, (m, 1))
 diffDataArray = (testDataArray - trainData) ** 2
 sumDataArray = diffDataArray.sum(axis=1) ** 0.5
 # 對結(jié)果進(jìn)行排序
 sumDataSortedArray = sumDataArray.argsort()
 
 classCount = {}
 for i in range(k):
  labelName = trainLabel[list(sumDataSortedArray).index(i)]
  classCount[labelName] = classCount.get(labelName, 0)+1
 classCount = sorted(classCount.items(), key=lambda x: x[1], reversed=True)
 return classCount[0][0]
 

# 數(shù)據(jù)測試
def datingTest(file):
 datingData, datingTrainData, datingTrainLabel = loadDatingData(file)
 normValuesData = autoNorm(datingTrainData)
 
 
 errorCount = 0
 ratio = 0.10
 total = datingTrainData.shape(0)
 numberTest = int(total * ratio)
 for i in range(numberTest):
  res = KNNClassifier(normValuesData[i], normValuesData[numberTest:m], datingTrainLabel, 5)
  if res != datingTrainLabel[i]:
   errorCount += 1
 print("The total error rate is : {}\n".format(error/float(numberTest)))

if __name__ == "__main__":
 FILEPATH = "./datingTestSet1.txt"
 datingTest(FILEPATH)
# python 第三方包實(shí)現(xiàn)
import pandas as pd
import numpy as np
from sklearn.neighbors import KNeighborsClassifier

if __name__ == "__main__":
 FILEPATH = "./datingTestSet1.txt"
 datingData, datingTrainData, datingTrainLabel = loadDatingData(FILEPATH)
 normValuesData = autoNorm(datingTrainData)
 errorCount = 0
 ratio = 0.10
 total = normValuesData.shape[0]
 numberTest = int(total * ratio)
 
 k = 5
 clf = KNeighborsClassifier(n_neighbors=k)
 clf.fit(normValuesData[numberTest:total], datingTrainLabel[numberTest:total])
 
 for i in range(numberTest):
  res = clf.predict(normValuesData[i].reshape(1, -1))
  if res != datingTrainLabel[i]:
   errorCount += 1
 print("The total error rate is : {}\n".format(errorCount/float(numberTest)))

以上就是python實(shí)現(xiàn)KNN近鄰算法的詳細(xì)內(nèi)容，更多關(guān)于python實(shí)現(xiàn)KNN近鄰算法的資料請關(guān)注腳本之家其它相關(guān)文章！

您可能感興趣的文章: