House Prices - Regression.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://s3.amazonaws.com/keras-datasets/boston_housing.npz\n",
      "57344/57026 [==============================] - 0s 6us/step\n"
     ]
    }
   ],
   "source": [
    "from keras.datasets import boston_housing\n",
    "(train_data, train_targets), (test_data, test_targets) = boston_housing.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "def normalize(x, xtest, axis = 0):\n",
    "    m = np.mean(x, axis = axis)\n",
    "    meanCentred = x-m\n",
    "    meanCentredTest = xtest - m\n",
    "    sigma = np.std(x, axis = axis)\n",
    "    return [meanCentred/sigma, meanCentredTest/sigma]\n",
    "[x,xTest] = normalize(train_data, test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import models\n",
    "from keras import layers\n",
    "def build_model():\n",
    "    model = models.Sequential()\n",
    "    model.add(layers.Dense(64, activation = 'relu', input_shape = (train_data.shape[1],)))\n",
    "    model.add(layers.Dense(64, activation = 'relu'))\n",
    "    model.add(layers.Dense(1))\n",
    "    model.compile(optimizer = 'rmsprop', loss = 'mse' , metrics = ['mae'])\n",
    "    return model\n",
    "\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "fold  0\n",
      "fold  1\n",
      "fold  2\n",
      "fold  3\n",
      "fold  4\n"
     ]
    }
   ],
   "source": [
    "k = 5\n",
    "num_val_samples = len(x)//k\n",
    "num_epochs = 200\n",
    "all_scores = []\n",
    "all_mae_histories = []\n",
    "for i in range(k):\n",
    "    print(\"fold \",i)\n",
    "    val_data = x[i*num_val_samples:(i+1)*num_val_samples]\n",
    "    val_targets = train_targets[i*num_val_samples:(i+1)*num_val_samples]\n",
    "    xt = np.concatenate([x[:i*num_val_samples], x[(i+1)*num_val_samples:]], axis = 0)\n",
    "    yt = np.concatenate([train_targets[:i*num_val_samples],train_targets[(i+1)*num_val_samples:]], axis=0)\n",
    "    model = build_model()\n",
    "    hist = model.fit(xt,yt, epochs = num_epochs, batch_size = 1, verbose = 0)\n",
    "    mae_history = hist.history['mean_absolute_error']\n",
    "    all_mae_histories.append(mae_history)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD8CAYAAABn919SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\noIuIhIQCXUQkJBToIiIh8f8BNOneZJYQNkAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x148c00b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "def smooth_curve(points, factor = 0.9):\n",
    "    smoothed_points = []\n",
    "    for pt in points:\n",
    "        if smoothed_points:\n",
    "            prev = smoothed_points[-1]\n",
    "            smoothed_points.append(prev*factor + pt*(1-factor))\n",
    "        else:\n",
    "            smoothed_points.append(pt)\n",
    "    return smoothed_points\n",
    "average_mae_hist = [np.mean([x[i] for x in all_mae_histories]) for i in range(num_epochs)]\n",
    "smooth_hist = smooth_curve(average_mae_hist)\n",
    "plt.plot(smooth_hist[:])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}