{ "cells": [ { "cell_type": "code", "execution_count": 37, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "'1.13.1'" ] }, "execution_count": 37, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import tensorflow as tf\n", "import tensorflow.examples.tutorials.mnist.input_data as input_data\n", "\n", "tf.__version__" ] }, { "cell_type": "code", "execution_count": 36, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Extracting MNIST_data/train-images-idx3-ubyte.gz\n", "Extracting MNIST_data/train-labels-idx1-ubyte.gz\n", "Extracting MNIST_data/t10k-images-idx3-ubyte.gz\n", "Extracting MNIST_data/t10k-labels-idx1-ubyte.gz\n" ] } ], "source": [ "MNIST=input_data.read_data_sets(\"MNIST_data\", one_hot=True)" ] }, { "cell_type": "code", "execution_count": 34, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "0.9149\n" ] } ], "source": [ "# 定义一个占位符x\n", "x = tf.placeholder(tf.float32, [None, 784]) # 张量的形状是[None, 784]，None表第一个维度任意\n", "\n", "# 定义变量W,b,是可以被修改的张量，用来存放机器学习模型参数\n", "W = tf.Variable(tf.zeros([784, 10]))\n", "b = tf.Variable(tf.zeros([10]))\n", "\n", "# 实现模型, y是预测分布\n", "y = tf.nn.softmax(tf.matmul(x, W) + b)\n", "\n", "# 训练模型，y_是实际分布\n", "y_ = tf.placeholder(\"float\", [None, 10])\n", "cross_entropy = -tf.reduce_sum(y_ * tf.log(y)) # 交叉嫡，cost function\n", "\n", "# 使用梯度下降来降低cost，学习速率为0.01\n", "train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)\n", "\n", "# 初始化已经创建的变量\n", "init = tf.global_variables_initializer()\n", "\n", "# 在一个Session中启动模型，并初始化变量\n", "sess = tf.Session()\n", "sess.run(init)\n", "\n", "# # 训练模型，运行1000次，每次随机抽取100个\n", "for i in range(1, 1000):\n", " batch_xs, batch_ys = MNIST.train.next_batch(100)\n", " sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})\n", "\n", "# 验证正确率\n", "correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))\n", "accuracy = tf.reduce_mean(tf.cast(correct_prediction, 'float'))\n", "print(sess.run(accuracy, feed_dict={x: MNIST.test.images, y_: MNIST.test.labels}))" ] }, { "cell_type": "code", "execution_count": 41, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "当前迭代次数为0，训练正确率为0.8999999761581421\n", "当前迭代次数为200，训练正确率为0.9200000166893005\n", "当前迭代次数为400，训练正确率为0.9100000262260437\n", "当前迭代次数为600，训练正确率为0.9399999976158142\n", "当前迭代次数为800，训练正确率为0.8999999761581421\n", "当前迭代次数为1000，训练正确率为0.8799999952316284\n", "当前迭代次数为1200，训练正确率为0.9200000166893005\n", "当前迭代次数为1400，训练正确率为0.9100000262260437\n", "当前迭代次数为1600，训练正确率为0.9200000166893005\n", "当前迭代次数为1800，训练正确率为0.8399999737739563\n", "当前迭代次数为2000，训练正确率为0.8899999856948853\n", "当前迭代次数为2200，训练正确率为0.8700000047683716\n", "当前迭代次数为2400，训练正确率为0.8999999761581421\n", "当前迭代次数为2600，训练正确率为0.8899999856948853\n", "当前迭代次数为2800，训练正确率为0.8999999761581421\n", "当前迭代次数为3000，训练正确率为0.9100000262260437\n", "当前迭代次数为3200，训练正确率为0.949999988079071\n", "当前迭代次数为3400，训练正确率为0.9300000071525574\n", "当前迭代次数为3600，训练正确率为0.8999999761581421\n", "当前迭代次数为3800，训练正确率为0.9100000262260437\n", "当前迭代次数为4000，训练正确率为0.9700000286102295\n", "当前迭代次数为4200，训练正确率为0.8899999856948853\n", "当前迭代次数为4400，训练正确率为0.8999999761581421\n", "当前迭代次数为4600，训练正确率为0.9300000071525574\n", "当前迭代次数为4800，训练正确率为0.9399999976158142\n", "当前迭代次数为5000，训练正确率为0.8899999856948853\n", "当前迭代次数为5200，训练正确率为0.9100000262260437\n", "当前迭代次数为5400，训练正确率为0.9100000262260437\n", "当前迭代次数为5600，训练正确率为0.9100000262260437\n", "当前迭代次数为5800，训练正确率为0.9300000071525574\n", "当前迭代次数为6000，训练正确率为0.9200000166893005\n", "当前迭代次数为6200，训练正确率为0.8700000047683716\n", "当前迭代次数为6400，训练正确率为0.9200000166893005\n", "当前迭代次数为6600，训练正确率为0.9399999976158142\n", "当前迭代次数为6800，训练正确率为0.8999999761581421\n", "当前迭代次数为7000，训练正确率为0.9300000071525574\n", "当前迭代次数为7200，训练正确率为0.9300000071525574\n", "当前迭代次数为7400，训练正确率为0.9300000071525574\n", "当前迭代次数为7600，训练正确率为0.8799999952316284\n", "当前迭代次数为7800，训练正确率为0.9100000262260437\n", "当前迭代次数为8000，训练正确率为0.8999999761581421\n", "当前迭代次数为8200，训练正确率为0.9599999785423279\n", "当前迭代次数为8400，训练正确率为0.9300000071525574\n", "当前迭代次数为8600，训练正确率为0.8799999952316284\n", "当前迭代次数为8800，训练正确率为0.9200000166893005\n", "当前迭代次数为9000，训练正确率为0.8600000143051147\n", "当前迭代次数为9200，训练正确率为0.9700000286102295\n", "当前迭代次数为9400，训练正确率为0.9200000166893005\n", "当前迭代次数为9600，训练正确率为0.8899999856948853\n", "当前迭代次数为9800，训练正确率为0.8799999952316284\n" ] } ], "source": [ " \n", "for i in range(10000):\n", " batch = MNIST.train.next_batch(100)\n", " if i % 200 == 0:\n", " print(\"当前迭代次数为{}，训练正确率为{}\".format(i, accuracy.eval(feed_dict={x:batch[0], y_: batch[1]},session=sess)))\n", " train_step.run(feed_dict={x: batch[0], y_: batch[1]},session=sess)" ] }, { "cell_type": "code", "execution_count": 45, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " 测试正确率为 0.912\n" ] } ], "source": [ "#测试阶段\n", "print(\" 测试正确率为 %g\"%accuracy.eval(feed_dict={x: MNIST.test.images, y_: MNIST.test.labels},session=sess)) " ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.3" }, "toc": { "base_numbering": 1, "nav_menu": {}, "number_sections": true, "sideBar": true, "skip_h1_title": false, "title_cell": "Table of Contents", "title_sidebar": "Contents", "toc_cell": false, "toc_position": {}, "toc_section_display": true, "toc_window_display": false } }, "nbformat": 4, "nbformat_minor": 2 }