机器学习Sklearn实战——KNN算法

scikit-learn

• Simple and efficient tools for predictive data analysis
• Accessible to everybody, and reusable in various contexts
• Built on NumPy, SciPy, and matplotlib
• Open source, commercially usable - BSD license

KNN算法入门

• k近邻算法：nearest neighbors classification
• k定义多少个邻居
• 物以类聚，人以群分
• 教会计算机根据邻居分类，邻居：距离比较近
• 
• 距离越小，越近，d<10规定邻居
• 计算——>数学公式——>计算结构
• 计算获取邻居（5个邻居）——>统计，类别（富有，贫穷），富有4个，贫穷1个
• 4>1 富有>贫穷
• 结论：判断出来，我也是富有，概率，概率低
• 人分类时，有可能出现错误

from sklearn.neighbors import KNeighborsClassifier
import pandas as pd
movie = pd.read_excel("/Users/zhucan/Desktop/movies.xlsx",sheet_name = 1)
data = movie.iloc[:,1:3]
target = movie['分类情况']
knn = KNeighborsClassifier(n_neighbors=5)
knn.fit(data,target)
X_test = pd.DataFrame({"武打镜头":[100,67,1],"接吻镜头":[3,2,10]})
print(knn.predict(X_test))

['动作片' '动作片' '爱情片']

KNN鸢尾花分类

import numpy as np
from sklearn.neighbors import KNeighborsClassifier
from sklearn import datasets
iris = datasets.load_iris()
x = iris["data"]
y = iris["target"]
index = np.arange(150)
np.random.shuffle(index)
x_train,x_test = x[index[:100]],x[index[100:]]
y_train,y_test = y[index[:100]],y[index[100:]]
knn = KNeighborsClassifier(n_neighbors=5)
knn.fit(x_train,y_train)
y_=knn.predict(x_test)
print(y_)
print("--------------------")
print(y_test)

[1 1 0 1 2 1 1 0 0 0 2 0 0 1 1 2 1 2 0 1 0 0 2 0 1 2 0 2 0 0 0 0 1 2 1 1 1
 0 0 0 0 1 2 2 2 1 1 2 1 1]
--------------------
[1 1 0 1 2 1 1 0 0 0 2 0 0 1 1 2 1 2 0 1 0 0 2 0 1 2 0 2 0 0 0 0 1 2 1 1 1
 0 0 0 0 1 1 2 2 1 1 2 1 1]

KNN算法调参数

print(knn.score(x_test,y_test))

0.96

proba_ = knn.predict_proba(x_test)
print(proba_)
print(proba_.argmax(axis=1))
print(y_)

[[0.  1.  0. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [1.  0.  0. ]
 [0.  1.  0. ]
 [0.  1.  0. ]
 [0.  0.  1. ]
 [0.  1.  0. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [0.  0.2 0.8]
 [1.  0.  0. ]
 [0.  1.  0. ]
 [0.  1.  0. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [0.  0.  1. ]
 [0.  0.  1. ]
 [1.  0.  0. ]
 [1.  0.  0. ]
 [0.  1.  0. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [0.  0.  1. ]
 [1.  0.  0. ]
 [0.  0.2 0.8]
 [0.  1.  0. ]
 [0.  0.2 0.8]
 [1.  0.  0. ]
 [1.  0.  0. ]
 [0.  1.  0. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [0.  1.  0. ]
 [0.  1.  0. ]
 [0.  0.  1. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [0.  0.  1. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [1.  0.  0. ]
 [1.  0.  0. ]
 [1.  0.  0. ]
 [0.  0.  1. ]
 [0.  1.  0. ]
 [1.  0.  0. ]
 [0.  1.  0. ]
 [0.  1.  0. ]
 [0.  0.  1. ]]
[1 0 2 0 1 1 2 1 0 2 2 0 1 1 0 2 2 2 0 0 1 0 2 2 0 2 1 2 0 0 1 0 2 1 1 2 0
 2 2 0 2 0 0 0 2 1 0 1 1 2]
[1 0 2 0 1 1 2 1 0 2 2 0 1 1 0 2 2 2 0 0 1 0 2 2 0 2 1 2 0 0 1 0 2 1 1 2 0
 2 2 0 2 0 0 0 2 1 0 1 1 2]

knn = KNeighborsClassifier(n_neighbors=5,weights="uniform"/"distance",p=1/2)
p=1曼哈顿距离
p=2欧氏距离

n_neighbors最好不要超过样本数量开平方

KNN手写数字识别

import numpy as np
from sklearn.neighbors import KNeighborsClassifier
import cv2
import matplotlib.pyplot as plt
%matplotlib inline

digit = cv2.imread("./data/0/0_101.bmp")
digit = cv2.cvtColor(digit,code=cv2.COLOR_BGR2GRAY)

X=[]
for i in range(10):
    for j in range(1,501):
        digit = cv2.imread("./data/%d/%d_%d.bmp"%(i,i,j))
        X.append(digit[:,:,0])

#数据X和目标值y是一一对应
X = np.asarray(X)
y = [i for i in range(10)]*500
y.sort()

index = np.random.randint(0,5000,size=1)[0]
digit = X[index]
print(y[index])

plt.imshow(digit,cmap=plt.cm.gray)

from sklearn.model_selection import train_test_split
X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2)
X_train = X_train.reshape(4000,-1)
knn = KNeighborsClassifier(n_neighbors=63)
knn.fit(X_train,y_train)
X_test = X_test.reshape(1000,-1)
y_ = knn.predict(X_test)
knn.score(X_test,y_test)

二值化操作
for i in range(5000):
    for y in range(28):
        for x in range(28):
            if X[i][y,x] < 200:
                X[i][y,x] = 0
            else:
                X[i][y, x] = 0
train_test_split(X,y,test_size=0.2,random_state=100)