notebooks/linear_regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "データの作成"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfMAAAFXCAYAAAC2gqezAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4a906afc90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# coding:utf-8\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "b_train = -1\n",
    "W_train = 0.7\n",
    "x_train = np.linspace(0, 1.0, 100)\n",
    "y_train = x_train * W_train + b_train\n",
    "\n",
    "plt.figure(1)\n",
    "plt.plot(x_train, y_train, 'ro', label='Data')\n",
    "plt.plot(x_train, y_train, 'k-', label='Line')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Step 0, cost 1.134326, weight 1.816771, bias -0.544337\n",
      "Step 100, cost 0.002665, weight 0.875718, bias -1.094173\n",
      "Step 200, cost 0.000179, weight 0.745550, bias -1.024412\n",
      "Step 300, cost 0.000012, weight 0.711807, bias -1.006328\n",
      "Step 400, cost 0.000001, weight 0.703061, bias -1.001640\n",
      "Step 500, cost 0.000000, weight 0.700793, bias -1.000425\n",
      "Step 600, cost 0.000000, weight 0.700206, bias -1.000110\n",
      "Step 700, cost 0.000000, weight 0.700054, bias -1.000029\n",
      "Step 800, cost 0.000000, weight 0.700015, bias -1.000008\n",
      "Step 900, cost 0.000000, weight 0.700004, bias -1.000002\n",
      "Step 1000, cost 0.000000, weight 0.700003, bias -1.000002\n",
      "Step 1100, cost 0.000000, weight 0.700003, bias -1.000002\n",
      "Step 1200, cost 0.000000, weight 0.700003, bias -1.000002\n",
      "Step 1300, cost 0.000000, weight 0.700003, bias -1.000002\n",
      "Step 1400, cost 0.000000, weight 0.700003, bias -1.000002\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "\n",
    "# 変数の定義\n",
    "x = tf.placeholder(tf.float32, name=\"input\")\n",
    "y = tf.placeholder(tf.float32, name=\"output\")\n",
    "w = tf.Variable(np.random.randn(), name=\"weight\")\n",
    "b = tf.Variable(np.random.randn(), name=\"bias\")\n",
    "\n",
    "# 線形回帰のモデル\n",
    "y_pred = tf.add(tf.multiply(x, w), b)\n",
    "\n",
    "# 損失関数\n",
    "loss = tf.reduce_mean(tf.pow(y_pred - y, 2))\n",
    "\n",
    "# Optimizer\n",
    "# 勾配降下法\n",
    "learning_rate = 0.1\n",
    "train_op = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss)\n",
    "\n",
    "# セッション\n",
    "with tf.Session() as sess:\n",
    "\n",
    "  # 変数の初期化\n",
    "  init_op = tf.global_variables_initializer()\n",
    "  sess.run(init_op)\n",
    "\n",
    "  # トレーニング回数\n",
    "  training_step = 1500\n",
    "  validation_step = 100\n",
    "  \n",
    "  #トレーニング\n",
    "  for step in range(training_step):\n",
    "    sess.run(train_op, feed_dict={x:x_train, y:y_train})\n",
    "\n",
    "    # 途中経過表示\n",
    "    if step % validation_step == 0:\n",
    "        loss_output = sess.run(loss, feed_dict={x:x_train, y:y_train})\n",
    "        w_output = sess.run(w)\n",
    "        b_output = sess.run(b)\n",
    "        print \"Step %i, cost %f, weight %f, bias %f\" % (step, loss_output, w_output, b_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}