1. K近邻算法
当有多种类别数据时,我们常常面临着对新加入的数据进行分类的问题,例如,根据口味和色泽划分新的葡萄酒的类别,根据内容形式划分新上映电影的类别,根据过往人脸特征进行人脸识别等。这些问题都可以采用机器学习中非常经典的K近邻算法来解决。本章将讲解K近邻算法的原理与代码实现,然后通过手写数字识别模型的案例来巩固所学知识。
2.K近邻算法基本原理
英文为K-Nearest Neighbor,因而又简称KNN算法
K近邻算法的原理非常简单:对于一个新样本,K近邻算法的目的就是在已有数据中寻找与它最相似的K个数据,或者说“离它最近”的K个数据,如果这K个数据大多数属于某个类别,则该样本也属于这个类别
3.手写数字识别的原理
手写数字识别,或者说图像识别的本质就是把如下图所示的一张图片转换成计算机能够处理的数字形式。
3.1. 图像二值化
如下图所示是将图片格式的数字8转换成由0和1组成的“新的数字8”。这是一个32×32的矩阵,数字1代表有颜色的地方,数字0代表无颜色的地方,这样就完成了手写数字识别的第一步也是最关键的一步:将图片转换为计算机能识别的内容——数字0和1
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
3.2. 二维数组转换为一维数组
(1, 1024)
4. 手写数字识别的代码实现
import pandas as pd
df = pd.read_excel('手写字体识别.xlsx')
print(df.head(10))
X = df.drop(columns=['对应数字'])
y= df['对应数字']
#from sklearn.preprocessing import StandardScaler
#X = StandardScaler.fit_transform(X)
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,random_state=123)
from sklearn.neighbors import KNeighborsClassifier as KNN
knn = KNN(n_neighbors=5)
knn.fit(X_train,y_train)
y_pred = knn.predict(X_test)
a = pd.DataFrame()
a['真实值']= list(y_test)
a['预测值'] = list(y_pred)
#print(a.head())
from sklearn.metrics import accuracy_score
score = accuracy_score(y_test,y_pred)
print(score)
# 调优
# from sklearn.model_selection import GridSearchCV
# parameters = {'n_neighbors':[1,2,4,5,6,7,8,9]}
# knn = KNN()
# grid_search = GridSearchCV(knn,parameters,cv=5) #5折交叉验证
# grid_search.fit(X_train,y_train)
# print(grid_search.best_params_['n_neighbors']) #获取参数最优值
from PIL import Image
img = Image.open('数字8.png')
img = img.resize((32,32))
# 灰度处理
img = img.convert('L')
#img.show()
import numpy as np
img_new = img.point(lambda x: 0 if x>128 else 1)
arr = np.array(img_new)
for i in range(arr.shape[0]):
print(arr[i])
arr_new = arr.reshape(1,-1)
print(arr_new.shape)
answer = knn.predict(arr_new)
print('图片中的数字为:'+str(answer))
图片中的数字为:[8]