【机器学习】【KNN】线性扫描算法，python实现识别手写数字的系统

最新推荐文章于 2024-02-22 20:19:26 发布
最新推荐文章于 2024-02-22 20:19:26 发布
阅读量377 收藏 2
点赞数
分类专栏： 人工智能 机器学习 跟我一起学机器学习 Machine Learning
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/u012421852/article/details/79618438
版权

1.样本集资源

1.1样本集所在目录情况


1.2训练样本集所在目录情况

训练样本集中每个文件命名规则介绍：

x_y.txt

x是文件中手写数字的真实数值

y是第y个手写数字值为x的样本集文件


1.3测试样本集所在目录情况

测试样本集中每个文件命名规则介绍：

x_y.txt

x是文件中手写数字的真实数值

y是第y个手写数字值为x的样本集文件


1.4样本集下载链接

https://download.csdn.net/download/u012421852/10316218

2.KNN算法识别手写数字的系统实现代码

'''
@author: 蔚蓝的天空Tom
Talk is cheap, show me the code
Aim:KNN(k Nearest Neighbors)算法实现手写数字的识别
'''
import numpy as np
import math
#import operator
from os import listdir


train_set_dir = 'D:\\tom\\data\\digits\\trainingDigits\\'
test_set_dir =  'D:\\tom\\data\\digits\\testDigits\\'


#开根号函数的向量化
def sqrt_func(x):
    func = np.frompyfunc(math.sqrt, 1, 1)
    return func(x)


#计算欧氏距离
def euclidean_distance(test_data, train_data):
    #训练数据总数
    count = train_data.shape[0]
    #计算[x1-x2, y1-y2]
    diff_a = np.tile(test_data, (count, 1)) - train_data
    #计算[(x1-x2)^2, (y1-y2)^2]
    square_dist_a = np.sum(np.square(diff_a), axis=1)
    #计算sqrt((x1-x2)^2, (y1-y2)^2)
    dist_a = sqrt_func(square_dist_a)
    return dist_a


#KNN classify
def knn_classify(test_data, train_data, train_label, k=3):  
    #计算欧氏距离
    dist = euclidean_distance(test_data, train_data)
    
    #依次返回数值从小到大元素的下标
    #如果dist:100,400,300,200,600,500
    #那么index:0,3,2,1,5,4
    index = np.argsort(dist)
    
    #用字典统计每个标签的投票总数
    #{'A':0, 'B':3}表示标签A投票0次，标签B投票3次
    label_cnt_dict={}
    for i in range(k):
        #获取第index[i]个样本所在的标签voteLabel
        voteLabel = train_label[index[i]]
        #对voteLabel标签的投票数统计加1
        label_cnt_dict[voteLabel] = label_cnt_dict.get(voteLabel,0) + 1
    
    #找到投票次数最多的标签
    max_cnt = 0
    for key,value in label_cnt_dict.items():
        if value > max_cnt:
            max_cnt = value
            winner = key


    #胜利者标签就是测试数据被预测的分类
    return winner


'''
手写数字二值化后处理后的0_1.txt文件内容格式如下所示：
00000000000001100000000000000000
00000000000011111100000000000000
00000000000111111111000000000000
00000000011111111111000000000000
00000001111111111111100000000000
00000000111111100011110000000000
00000001111110000001110000000000
00000001111110000001110000000000
00000011111100000001110000000000
00000011111100000001111000000000
00000011111100000000011100000000
00000011111100000000011100000000
00000011111000000000001110000000
00000011111000000000001110000000
00000001111100000000000111000000
00000001111100000000000111000000
00000001111100000000000111000000
00000011111000000000000111000000
00000011111000000000000111000000
00000000111100000000000011100000
00000000111100000000000111100000
00000000111100000000000111100000
00000000111100000000001111100000
00000000011110000000000111110000
00000000011111000000001111100000
00000000011111000000011111100000
00000000011111000000111111000000
00000000011111100011111111000000
00000000000111111111111110000000
00000000000111111111111100000000
00000000000011111111110000000000
00000000000000111110000000000000
'''
def imgTxt_2_array(file_name):
    ret_img_a = np.zeros((1,1024)) #32 * 32 = 1024
    f = open(file_name)
    line_list = f.readlines()
    line_cnt = len(line_list)
    cnt = 0
    for i in range(line_cnt):
        line_str = line_list[i]
        #len()函数把'\n'也计算在内，真是字符总数要减去1
        line_str_cnt = len(line_str) - 1 
        for j in range(line_str_cnt):
            ret_img_a[0, cnt+j] = int(line_str[j])
        cnt += line_str_cnt
    return ret_img_a


#获取0_1.txt中手写数字的真实数字0(根据文件名结构获取)
def get_real_num(file_name_str = '0_1.txt'):
    #以为分隔符，得到子字符串
    str_list = file_name_str.split('.') #['0_1', 'txt']
    #取出包含num的子字符串'0_1'
    name_str = str_list[0]
    #以_为分隔符，得到子字符串
    int_list = name_str.split('_') #['0', '1']
    #取出第一个数字字符串，就是真实的手写数字
    real_num_str = int_list[0]
    return int(real_num_str)


def knn_classify_handwrite_number():
    #训练样本集所有文件名
    train_file_list = listdir(train_set_dir)#['0_1.txt', '0_2.txt',……]
    #训练样本集文件总数
    train_file_cnt = len(train_file_list)
    #训练样本集初始化
    train_file_a = np.zeros((train_file_cnt,1024))
    #训练标签初始化
    train_labels = []
    #获取训练样本集和训练标签
    for i in range(train_file_cnt):
        file_name_str = train_file_list[i] #'0_1.txt'
        train_labels.append(get_real_num(file_name_str))
        train_file_a[i,:] = imgTxt_2_array(train_set_dir+file_name_str)
        
    #测试样本集所有文件名
    test_file_list = listdir(test_set_dir)#['0_1.txt', '0_2.txt',……]
    #测试样本集文件总睡
    test_file_cnt = len(test_file_list)
    #测试分类错误总数
    err_cnt = 0.0
    #逐个对测试样本集文件进行数字识别
    for i in range(test_file_cnt):
        file_name_str = test_file_list[i]#'0_1.txt'
        test_img_a = imgTxt_2_array(test_set_dir+file_name_str)
        test_result = knn_classify(test_img_a, train_file_a, train_labels, 3)
        if (test_result != get_real_num(file_name_str)):
            err_cnt += 1.0
        print("KNN result: %d, real answer: %d" % (test_result, get_real_num(file_name_str)))
    #错误率
    err_rate = err_cnt / float(test_file_cnt)
    #正确率
    correct_rate = 1 - err_rate
    return err_cnt, err_rate, correct_rate
    
if __name__=='__main__':
    err_cnt, err_rate, correct_rate = knn_classify_handwrite_number()
    print("\nKNN error count  is: %d" % err_cnt)
    print("\nKNN error rate   is: %f" % err_rate)
    print('\nKNN correct rate is: %f' % correct_rate)
    

3.识别结果

3.1中间变量结果

3.2运行日志结果

KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 4, real answer: 4
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 3, real answer: 5
KNN result: 6, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 5, real answer: 5
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 6, real answer: 6
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 7, real answer: 7
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 6, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 3, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 1, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 1, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 8, real answer: 8
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 1, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 7, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9
KNN result: 9, real answer: 9

KNN error count  is: 11

KNN error rate   is: 0.011628

KNN correct rate is: 0.988372

4.参考文档

1.【机器学习】k-近邻算法（kNN） 学习笔记：https://blog.csdn.net/fuxuemingzhu/article/details/68982043

2.机器学习（一）——K-近邻（KNN）算法：http://www.cnblogs.com/ybjourney/p/4702562.html

enjoy it~

(end)

CV_ML_DP
关注
  • 0
    点赞
  • 踩
  • 2
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
专栏目录
KNN算法实现手写数字识别的三种方法（源码）
04-27
KNN算法使用MNIST数据集、0-1二值图集、自写的数字图片集，做手写数字识别的代码，文件夹分类明确。
机器学习(八)-基于KNN分类算法的手写识别系统
gf_lvah的博客
04-17 728
1 项目介绍 基于k-近邻分类器(KNN)的手写识别系统, 这里构造的系统只能识别数字0到9。 数据集和项目源代码 难点： 图形信息如何处理? 图像转换为文本格式 2 准备数据 将图像转换为测试向量 训练集： 目录trainingDigits 大约2000个例子 每个数字大约有200个样本; 测试集 目录testDigits 大约900个测试数据。 将图像格式化处理为一个向...
基于KNN算法的手写体数字识别
secretboys的博客
12-09 2942
基于KNN算法的手写体数字识别 KNN分类算法是一种经典的分类算法，属于懒惰学习算法的一种。 1.算法原理 工作原理：存在一个样本数据集合，也称作训练样本集，并且样本集中每个数据都存在标签，即我们知道样本集中每一数据与所属分类的对应关系。输入没有标签的新数据后，将新数据的每个特征与样本集中数据对应的特征进行比较，然后算法提取样本集中特征最相似数据（最近邻）的分类。一般来说，我们只选择样本数据集中前k个最相似的数据，这就是kNN算法中k的出处，通常k是不大于20的整数。最后，选择k个最相似数据中出现次数最多的
机器学习实战笔记——基于KNN算法的手写识别系统
henryzhihua
01-17 4645
利用k-近邻分类器实现手写识别系统，训练数据集大约2000个样本，每个数字大约有200个样本，每个样本保存在一个txt文件中，手写体图像本身是32X32的二值图像，如下图所示： 首先，我们需要将图像格式化处理为一个向量，把一个32X32的二进制图像矩阵通过img2vector()函数转换为1X1024的向量： def img2vector(filename): retu
机器学习--knn手写数字识别系统
XuShuai
05-07 3026
0. 刚接触java，并且在学习机器学习的相关算法，knn又非常的易于实现，于是就有了这个小系统。 1.knn算法简介： 2.该程序的功能主要有如下几个，       功能1:可以在面板上手写输入数字       功能2:可以对特定的区域进行截屏，因为要获取用户手写的数字，保存为图像，然后使用算法进行分析       功能3:可以对图片进行缩放，要保证图片的大小（维度）要和
机器学习算法Python实现.rar
最新发布
03-03
这个名为“机器学习算法Python实现”的压缩包文件，很可能是包含了一系列教程、代码示例或者完整的项目，旨在帮助用户理解并掌握如何在Python环境下实施各种机器学习算法。 Python中的机器学习库如Scikit-Learn、...
使用KNN算法与神经网络识别手写数字的程序。.zip
12-28
在本项目中，我们主要探讨的是两种机器学习算法——K近邻（K-Nearest Neighbors，简称KNN）和神经网络（Neural Networks），在手写数字识别任务中的应用。手写数字识别是计算机视觉领域的一个经典问题，通常用MNIST...
K邻近算法_K邻近算法；_机器学习；_python；_K._
10-03
**Python**作为数据科学和机器学习领域的主要编程语言，提供了多种库来实现KNN算法，其中最常用的是`sklearn`库。`sklearn`的`neighbors`模块包含了`KNeighborsClassifier`和`KNeighborsRegressor`两个类，分别用于...
手写数字识别
01-06
在这个项目中，我们将探讨如何使用Python编程语言以及两种不同的机器学习算法：支持向量机（SVM）和K近邻（KNN）来实现这一功能。 首先，我们要了解用于手写数字识别的数据集。最常用的是MNIST数据集，它包含60,000...
kNN.zip_0-9手势_40_knn识别_python 决策树·_手势
09-20
综上所述，这个项目结合了kNN算法和决策树，利用Python编程语言实现了手势识别，特别是手写数字的识别。通过训练和测试数据集，我们可以构建并评估一个高效的模型，能够在限定时间内完成识别任务。在实际应用中，...
手写体数字识别(Python+TensorFlow)-机器学习代码类资源(txt为微云链接)
07-16
使用Python+TensorFlow，全连接的神经网络，基于MNIST数据集。数据集包含60000张训练图片，10000张测试图片。MNIST_model文件夹是已经训练30000次的模型，也可以自己再训练。app.py文件可以测试自己的图片。
机器学习实例代码-手写数字识别.ipynb
08-10
练一个简单的全连接神经网络，即多层感知机（MLP），实战Python代码，下载即可使用，练一个简单的全连接神经网络，即多层感知机（MLP），实战Python代码，下载即可使用，
用 Python 实现一组手写数字识别.zip
04-20
用 Python 实现一组手写数字识别。使用keras+opencv进行简单的实现。首先进行图像中数字的目标检测与分割，将图片中的数字分离出来然后进行单独识别。使用的数据集为mnist手写数字识别库，采用卷积神经网络进行识别
Python手写数字识别实例带训练图集
10-08
PythonOpenCV用HOG+SVM识别手写数字实例程序。采用附件的图作为训练与测试集。
机器学习-神经网络-手写数字识别-KNN（有代码，有解析，有数据集，可视化）
m0_63669388的博客
03-07 3180
本文主要实现了基于tensorflow的手写数字识别，同时也有一个与KNN实现手写数字识别，内容较为细致，适合初步学习神经网络的友友。
Knn算法之手写识别系统
qq_43353213的博客
09-29 387
knn值之手写识别系统 导入包 import numpy as np from os import listdir import operator 介绍 1.os模块 os.listdir() 方法用于返回指定的文件夹包含的文件或文件夹的名字的列表。它不包括 . 和 … 即使它在文件夹中。 listdir()方法语法格式如下：os.listdir(path) path – 需要列出的目录路径 2.operator模块 本模块主要包括一些Python内部操作符对应的函数。这些函数主要分为几类：对
python实现基于KNN算法的手写数字识别系统 非常详细！！！！
springhammer的博客
04-25 2万+
可爱聪明善良的你关注有惊喜哦 实验知识点： 1、K 近邻分类算法 2、从文本文件中解析和导入数据 3、使用 Matplotlib 创建扩散图 4、归一化数值 首先介绍下监督学习和无监督学习： 监督学习：（数据有输入有输出），通过已有的一部分数据和输出数据之间的相应关系，生成一个函数，将输入映射到合适的输出，在准确的说有监督学习就是训练样本中必须有标签或者标记； 无监督学习：（数据只有输入），给定...
使用k-近邻算法构建手写识别系统（kNN）
热门推荐
兔子王
02-22 2万+
使用kNN算法构建手写识别系统。
课程设计(毕业设计)—基于机器学习KNN算法手写数字识别系统—计算机专业课程设计(毕业设计)
西瓜的博客
01-28 1万+
机器学习KNN算法手写数字识别系统 下载本文手写数字识别系统代码和课设报告的连接： 摘要 本文手写体数字识别系统的工作主要是运用K最邻近算法实现了对手写体数字的识别，支持上传本地图片和调用摄像头进行拍摄并截图识别两种识别的途径，同时有添加完善数据集、查看测试集的识别率的功能，形成了一个比较完整的手写数字识别系统。本文还运用python的GUI编程中的tkinter模块设计了一个简洁友好的用户界面。本文重点阐述了手写数字识别图像处理流程，运用KNN算法进行分类识别，同时运用数理统计的方法对K值的选取进行优
KNN算法实现数字识别系统
总结，这个文档提供的基于KNN的数字识别系统代码大全为读者提供了一个完整的实现流程，涵盖了从数据预处理、特征提取到模型训练和测试的各个阶段，对于学习和理解KNN算法及其在手写数字识别中的应用具有很高的参考...
评论
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值