机器学习Sklearn实战——KNN算法

scikit-learn Simple and efficient tools for predictive data analysisAccessible to everybody, and reusable in various contextsBuilt on NumPy, SciPy, and matplotlibOpen source, commercially usable - BSD license KNN算法入门 k近邻算法 nearest neighbors classificationk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)
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)