tensorflow really convenient, it appears that the depth of the learning needs of how to use the framework, how to model ~
1 ''' 2 softmax classifier for mnist 3 4 created on 2019.9.28 5 author: vince 6 ''' 7 import math 8 import logging 9 import numpy 10 import random 11 import matplotlib.pyplot as plt 12 import tensorflow as tf 13 from tensorflow.contrib.learn.python.learn.datasets.mnist import read_data_sets 14 from sklearn.metrics import accuracy_score 15 16 DEF main (): . 17 logging.basicConfig (Level = logging.info, 18 is the format = ' % (the asctime) S% (filename) S [Line:% (lineno) D]% (levelname) S% (Message) S ' , . 19 datefmt = ' % A,% B% D% the Y% H:% M:% S ' ); 20 is 21 is logging.info ( " Trainning the begin. " ); 22 is 23 is MNIST = read_data_sets ( ' ../data / MNIST ' , one_hot = True) # MNIST_data refers to the data file storage folder path, one_hot = True is encoded in the label using the one_hot 24 25 x = tf.placeholder(tf.float32, [None, 784]); 26 w = tf.Variable(tf.zeros([784, 10])); 27 b = tf.Variable(tf.zeros([10])); 28 y = tf.matmul(x, w) + b; 29 30 y_ = tf.placeholder(tf.float32, [None, 10]); 31 32 cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits = y, labels = y_)); 33 train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy); 34 35 sess = tf.InteractiveSession(); 36 tf.global_variables_initializer().run(); 37 for _ in range(1000): 38 batch_xs, batch_ys = mnist.train.next_batch(100); 39 sess.run(train_step, feed_dict = {x : batch_xs, y_ : batch_ys}); 40 41 logging.info("trainning end."); 42 logging.info("testing begin."); 43 44 correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1)); 45 accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)); 46 print(sess.run(accuracy, feed_dict = {x : mnist.test.images, y_:mnist.test.labels})); 47 48 logging.info("testing end."); 49 50 if __name__ == "__main__": 51 main();