K near algorithm (kNeighbrClassifier / KNN): principles of Euclidean distance + + recent vote (weight) + probability
Classified according distantly
Euclidean distance: the distance between each point in a multidimensional space
Disadvantages: time complexity and space complexity greater
Note meaning: when less training time sample data, be sure to sample the same proportion
KNN classification algorithm movie
Import numpy
Import PANDAS # Excel file is imported
from sklearn.neighbors Import KNeighborsClassifier # machine learning algorithm library without depth learning algorithm
Movie = pandas.read_excel (R & lt " D: \ the Python \ Code \ Machine-Learn \ 1-KNN \ data \ movie .xlsx " , sheet_name = 0)
Movie
| Movie name | Martial arts scene | Kissing scene | Classification | |
|---|---|---|---|---|
| 0 | Westward Journey | 36 | 1 | Action movie |
| 1 | SPL | 43 | 2 | Action movie |
| 2 | Former 3 | 0 | 10 | Romance |
| 3 | Wolf 2 | 59 | 1 | Action movie |
| 4 | Titanic | 1 | 15 | Romance |
| 5 | Xinyu wish | 2 | 19 | Romance |
= pandas.read_excel Movie (R & lt " D: \ the Python \ Code \-Learn Machine \-the KNN. 1 \ Data \ movie.xlsx " , SHEET_NAME = 0)
X = Movie [[ " martial arts Lens " , " kissing Lens " ]] # remove the training data in the training data
Y = movie [ " classification " ] # remove target
KNN = KNeighborsClassifier (= N_NEIGHBORS. 5 )
knn.fit (X, Y) # training data
# forecast movie "Speed" martial arts camera 50, kissing lens 2
x_text = pandas.DataFrame ({ " martial arts lens " : [50,3], " kissing lens " : [2,50]})
x_text
| Martial arts scene | Kissing scene | |
|---|---|---|
| 0 | 50 | 2 |
| 1 | 3 | 50 |
= get_result knn.predict (x_text)
get_proba = knn.predict_proba (x_text)
Print ( " Probability: {} " .format (get_proba))
Print ( " classification result: {} " .format (get_result))
Probability: [[0.6 0.4] [0.4 0.6]] results: [ 'action movie' 'love stories']
Film Classification operating principle
s=((movie["武打镜头"]-50)**2+(movie["接吻镜头"]-2)**2)**0.5 #根据knn算法求距离
index=s.sort_values().index #先将数据排序然后取出索引
fljg=movie["分类情况"][index[:5]]
print("\n{}".format(index),"\n分类：\n{}".format(fljg))
Int64Index([1, 3, 0, 2, 4, 5], dtype='int64') 分类： 1 动作片 3 动作片 0 动作片 2 爱情片 4 爱情片 Name: 分类情况, dtype: object
识别梵文
import numpy
import os
import matplotlib.pyplot as plt
from sklearn.neighbors import KNeighborsClassifier
img=plt.imread(r"D:\Python\代码\Machine-Learn\1-KNN\data\手写字母测试与训练\梵文识别学习\Test\character_1_ka\1339.png")
plt.imshow(img,cmap=plt.cm.gray)
<matplotlib.image.AxesImage at 0x1af31dcc048>
#将读取的图片保存到数组data中
def img_read(dir_name,data):
for filename in os.listdir(dir_name):
img=plt.imread(dir_name+"\\"+filename)
data.append(img)
def readTain(): #获取训练数据
data=[]
dir_path=r"D:\Python\代码\Machine-Learn\1-KNN\data\手写字母测试与训练\梵文识别学习\Train"
for dir_name in os.listdir(dir_path)[36:]:
dir_name=dir_path+"\\"+dir_name
img_read(dir_name,data)
return data
def readTest(): #获取测试数据
data_test=[]
dir_test_path=r"D:\Python\代码\Machine-Learn\1-KNN\data\手写字母测试与训练\梵文识别学习\Test"
for dir_name in os.listdir(dir_test_path)[36:]:
dir_name=dir_test_path+"\\"+dir_name
img_read(dir_name,data_test)
return data_test
x=readTain() #训练数据
train_x=numpy.array(x)
train_x_shape={}
train_x_shape["图片数量"]=test_x.shape[0]
train_x_shape["宽度"]=str(test_x.shape[1])+"px"
train_x_shape["高度"]=str(test_x.shape[2])+"px"
train_x_shape