具體knn算法概念參考knn代碼python實現
上面是參考《機器學習實戰》的代碼，和knn的思想

# _*_ encoding=utf8 _*_

import numpy as np
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

# 導入手寫體識別的數據
mnist = input_data.read_data_sets("../data", one_hot=True)

# 訓練集和測試集
X_train, Y_train = mnist.train.next_batch(5000) # 數據和labels
X_test, Y_test = mnist.test.next_batch(100)

# 定義輸入
x_train = tf.placeholder(tf.float32, shape=(None,784))
x_test = tf.placeholder(tf.float32, shape=(784))

# L1距離也就是城市街區距離 |x1-x2|+|y1-y2|
distance = tf.reduce_sum(tf.abs(www.leyou1178.cn/ tf.add(x_train,tf.negative(x_test))),reduction_indices = 1)

# 返回最近的坐標，0縱軸 1橫軸
pred = tf.arg_min(distance, 0)

accuracy = 0

# 初始化
init = tf.global_variables_initializer()

with tf.Session() as sess:

sess.run(init)

for i in range(len(X_test)):
# 獲取當前樣本的最近鄰索引，當前樣本和每一個訓練的樣本找一個最近的l1距離，得到這個最小距離的下標
nn_index = sess.run(pred, feed_dict={x_train:X_train, x_test: X_test[i, :]})

# 由最鄰近索引找到label,然後最鄰近的label與真實標簽比較 np.argmax找最大的下標
# 由l1距離找到的最小值對應的坐標，通過該最坐標找到對應行label的最大值的下標，這個下標對應的就是數字的大小
print("預測次數", i, "預測標簽:", np.argmax(Y_train[nn_index]),"真實標簽:", np.argmax(Y_test[i]))
# 計算準確率
if np.argmax(Y_www.xinghenyule.com train[nn_index]) www.089188.cn/==www.tianzunyule178.com np.argmax(Y_test[i]):
accuracy += 1

print("Accuracy:", float(accuracy)/len(X_test))

具體knn算法概念參考knn代碼python實現