一看就懂的Tensorflow实战（最近邻）
最近邻算法简介
k近邻模型的核心就是使用一种距离度量，获得距离目标点最近的k个点，根据分类决策规则，决定目标点的分类。[2]
距离度量(L1范数)：
K值选择：这里k为10。
分类决策规则：k近邻的分类决策规则是最为常见的简单多数规则，也就是在最近的K个点中，哪个标签数目最多，就把目标点的标签归于哪一类。
Tensorflow 最近邻
import numpy as np
import tensorflow as tf导入 mnist数据集
# Import MINST data
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("./data/", one_hot=True)Extracting ./data/train-images-idx3-ubyte.gz
Extracting ./data/train-labels-idx1-ubyte.gz
Extracting ./data/t10k-images-idx3-ubyte.gz
Extracting ./data/t10k-labels-idx1-ubyte.gz构建模型
# In this example, we limit mnist data
Xtr, Ytr = mnist.train.next_batch(5000) #5000 for training (nn candidates)
Xte, Yte = mnist.test.next_batch(10) #10 for testing
# tf Graph Input
xtr = tf.placeholder("float", [None, 784])
xte = tf.placeholder("float", [784])
# Nearest Neighbor calculation using L1 Distance
# Calculate L1 Distance
distance = tf.reduce_sum(tf.abs(tf.add(xtr, tf.negative(xte))), reduction_indices=1)
# Prediction: Get min distance index (Nearest neighbor)
pred = tf.argmin(distance, 0)补充：Tenosrflow中基本算术运算函数:[1]
- tf.add(x,y,name=None) #求和运算
- tf.subtract(x,y,name=None) #减法运算
- tf.multiply(x,y,name=None) #乘法运算
- tf.div(x,y,name=None) #除法运算
- tf.mod(x,y,name=None) #取模运算
- tf.abs(x,name=None)＃求绝对值
- tf.negative(x,name=None) ＃取负运算（ｙ＝－ｘ）
- tf.sign(x,name=None) ＃返回符合 ｘ大于０，则返回１，小于０，则返回－１
- tf.reciprocal(x,name=None) ＃取反运算
- tf.square(x,name=None) ＃计算平方
- tf.round(x,name=None) ＃舍入最接近的整数
- tf.pow(x,y,name=None) ＃幂次方
训练
accuracy = 0.
# Initialize the variables (i.e. assign their default value)
init = tf.global_variables_initializer()
# Start training
with tf.Session() as sess:
sess.run(init)
# loop over test data
for i in range(len(Xte)):
# Get nearest neighbor
# 5000个样本点分别和10个测试点计算距离
nn_index = sess.run(pred, feed_dict={xtr: Xtr, xte: Xte[i, :]})
print(nn_index)
# Get nearest neighbor class label and compare it to its true label
print ("Test", i, "Prediction:", np.argmax(Ytr[nn_index]), \
"True Class:", np.argmax(Yte[i]))
# Calculate accuracy
if np.argmax(Ytr[nn_index]) == np.argmax(Yte[i]):
accuracy += 1./len(Xte)
print ("Done!")
print ("Accuracy:", accuracy)190
Test 0 Prediction: 9 True Class: 9
475
Test 1 Prediction: 5 True Class: 5
3152
Test 2 Prediction: 7 True Class: 7
2413
Test 3 Prediction: 2 True Class: 2
1088
Test 4 Prediction: 2 True Class: 2
1427
Test 5 Prediction: 2 True Class: 2
4743
Test 6 Prediction: 7 True Class: 7
4826
Test 7 Prediction: 6 True Class: 6
4099
Test 8 Prediction: 5 True Class: 5
2421
Test 9 Prediction: 5 True Class: 5
Done!
Accuracy: 0.9999999999999999参考
[1] Tenosrflow中基本算术运算函数:https://blog.csdn.net/zSean/article/details/75097937
[2] 统计学习方法——K近邻模型:https://blog.csdn.net/qq_35082030/article/details/60965320
本文参与 腾讯云自媒体分享计划，分享自微信公众号。
原始发表：2018-04-10，如有侵权请联系 cloudcommunity@tencent.com 删除
评论
登录后参与评论
推荐阅读
目录