梯度下降实现一元一次线性回归(1).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x18283c88358>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "####### >>>>>填写<<<<<   设置你想要的真实斜率和截距， 以及数据点(50-200) >>>>>填写<<<<< ############\n",
    "true_w = 5\n",
    "true_b = 3\n",
    "num_samples = 200\n",
    "\n",
    "############################################################################################\n",
    "\n",
    "\n",
    "# 初始化随机数据\n",
    "X = tf.random.normal(shape=[num_samples, 1]).numpy()\n",
    "noise = tf.random.normal(shape=[num_samples, 1]).numpy()\n",
    "\n",
    "####### >>>>>填写<<<<< 在下述函数尾端，添加噪音项, 注意删除下划线 >>>>>填写<<<<<  ################################\n",
    "y = X * true_w + true_b + noise\n",
    "#######################################\n",
    "\n",
    "plt.scatter(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "W = tf.Variable(tf.random.uniform([1]))  # 随机初始化参数\n",
    "b = tf.Variable(tf.random.uniform([1]))\n",
    "\n",
    "# >>>>>填写<<<<<  定义一个随机一元一次线性函数  y = W * x + b   >>>>>填写<<<<<  ##\n",
    "# 思考 这里的x为什么是小写？ #\n",
    "def random_line(x):\n",
    " \n",
    "    y = W * x + b \n",
    "    return y\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x18283856b38>]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X, y)\n",
    "plt.plot(X, random_line(X), c='r')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def loss_fn( x, y):\n",
    "    y_ = random_line(x)\n",
    "    return tf.reduce_mean(tf.square(y_ - y))\n",
    "#交叉熵"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [0/50], loss [32.170], W is [0.193], b is [0.199]\n",
      "Epoch [1/50], loss [30.978], W is [0.284], b is [0.255]\n",
      "Epoch [2/50], loss [29.830], W is [0.374], b is [0.311]\n",
      "Epoch [3/50], loss [28.726], W is [0.462], b is [0.365]\n",
      "Epoch [4/50], loss [27.665], W is [0.549], b is [0.418]\n",
      "Epoch [5/50], loss [26.644], W is [0.634], b is [0.470]\n",
      "Epoch [6/50], loss [25.661], W is [0.717], b is [0.521]\n",
      "Epoch [7/50], loss [24.717], W is [0.799], b is [0.571]\n",
      "Epoch [8/50], loss [23.808], W is [0.879], b is [0.620]\n",
      "Epoch [9/50], loss [22.934], W is [0.957], b is [0.668]\n",
      "Epoch [10/50], loss [22.093], W is [1.035], b is [0.715]\n",
      "Epoch [11/50], loss [21.285], W is [1.110], b is [0.761]\n",
      "Epoch [12/50], loss [20.507], W is [1.185], b is [0.807]\n",
      "Epoch [13/50], loss [19.759], W is [1.257], b is [0.851]\n",
      "Epoch [14/50], loss [19.039], W is [1.329], b is [0.894]\n",
      "Epoch [15/50], loss [18.347], W is [1.399], b is [0.937]\n",
      "Epoch [16/50], loss [17.681], W is [1.468], b is [0.979]\n",
      "Epoch [17/50], loss [17.041], W is [1.535], b is [1.019]\n",
      "Epoch [18/50], loss [16.425], W is [1.601], b is [1.059]\n",
      "Epoch [19/50], loss [15.833], W is [1.666], b is [1.099]\n",
      "Epoch [20/50], loss [15.263], W is [1.730], b is [1.137]\n",
      "Epoch [21/50], loss [14.715], W is [1.792], b is [1.175]\n",
      "Epoch [22/50], loss [14.187], W is [1.854], b is [1.212]\n",
      "Epoch [23/50], loss [13.680], W is [1.914], b is [1.248]\n",
      "Epoch [24/50], loss [13.193], W is [1.973], b is [1.283]\n",
      "Epoch [25/50], loss [12.723], W is [2.031], b is [1.318]\n",
      "Epoch [26/50], loss [12.272], W is [2.087], b is [1.352]\n",
      "Epoch [27/50], loss [11.838], W is [2.143], b is [1.385]\n",
      "Epoch [28/50], loss [11.420], W is [2.198], b is [1.418]\n",
      "Epoch [29/50], loss [11.019], W is [2.251], b is [1.450]\n",
      "Epoch [30/50], loss [10.632], W is [2.304], b is [1.482]\n",
      "Epoch [31/50], loss [10.261], W is [2.356], b is [1.512]\n",
      "Epoch [32/50], loss [9.903], W is [2.406], b is [1.542]\n",
      "Epoch [33/50], loss [9.560], W is [2.456], b is [1.572]\n",
      "Epoch [34/50], loss [9.229], W is [2.505], b is [1.601]\n",
      "Epoch [35/50], loss [8.911], W is [2.552], b is [1.629]\n",
      "Epoch [36/50], loss [8.605], W is [2.599], b is [1.657]\n",
      "Epoch [37/50], loss [8.310], W is [2.645], b is [1.684]\n",
      "Epoch [38/50], loss [8.027], W is [2.690], b is [1.711]\n",
      "Epoch [39/50], loss [7.755], W is [2.735], b is [1.737]\n",
      "Epoch [40/50], loss [7.493], W is [2.778], b is [1.763]\n",
      "Epoch [41/50], loss [7.241], W is [2.821], b is [1.788]\n",
      "Epoch [42/50], loss [6.999], W is [2.863], b is [1.813]\n",
      "Epoch [43/50], loss [6.766], W is [2.904], b is [1.837]\n",
      "Epoch [44/50], loss [6.542], W is [2.944], b is [1.860]\n",
      "Epoch [45/50], loss [6.326], W is [2.983], b is [1.884]\n",
      "Epoch [46/50], loss [6.118], W is [3.022], b is [1.906]\n",
      "Epoch [47/50], loss [5.919], W is [3.060], b is [1.929]\n",
      "Epoch [48/50], loss [5.727], W is [3.097], b is [1.950]\n",
      "Epoch [49/50], loss [5.542], W is [3.134], b is [1.972]\n"
     ]
    }
   ],
   "source": [
    "### >>>>>填写<<<<< 通过改变epochs 的值，推荐起始10， 和learning rate 学习率 推荐0.1起始 观察梯度下降学习的线性函数w b的值，以及loss函数的变化>>>>>填写<<<<<  ######\n",
    "\n",
    "EPOCHS = 50\n",
    "LEARNING_RATE = 0.01\n",
    "#######################################################################################################################################\n",
    "\n",
    "\n",
    "\n",
    "for epoch in range(EPOCHS):  # 迭代次数\n",
    "    with tf.GradientTape() as tape:  # 追踪梯度\n",
    "        loss = loss_fn(X, y)  # 计算损失\n",
    "    dW, db = tape.gradient(loss, [W, b])  # 计算梯度\n",
    "    W.assign_sub(LEARNING_RATE * dW)  # 更新梯度\n",
    "    b.assign_sub(LEARNING_RATE * db)\n",
    "    # 输出计算过程\n",
    "    print('Epoch [{}/{}], loss [{:.3f}], W is [{:.3f}], b is [{:.3f}]'.format(epoch, EPOCHS, loss,\n",
    "                                                                     float(W.numpy()),\n",
    "                                                                     float(b.numpy())))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}