Keras 基础.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:35:38.376692Z",
     "start_time": "2019-01-11T12:35:38.339654Z"
    }
   },
   "outputs": [],
   "source": [
    "from keras import models \n",
    "from keras import layers\n",
    "\n",
    "net = models.Sequential() \n",
    "net.add(layers.Dense(512, activation='relu', input_shape=(28 * 28,)))\n",
    "net.add(layers.Dense(10, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:35:38.834279Z",
     "start_time": "2019-01-11T12:35:38.791280Z"
    }
   },
   "outputs": [],
   "source": [
    "net.compile(\n",
    "    optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:35:39.366430Z",
     "start_time": "2019-01-11T12:35:39.030413Z"
    }
   },
   "outputs": [],
   "source": [
    "from keras.datasets import mnist \n",
    " \n",
    "(train_images, train_labels), (test_images, test_labels) = mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:35:39.774384Z",
     "start_time": "2019-01-11T12:35:39.383385Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAADNJJREFUeJzt3WGoXPWZx/HfL9kWNA0aydWGNPF2iywr6qbrGBZcF5disFKMhVQasKRYen1RL1voi5W8qW8WZVnb9cUSuF1DI7Q21TQxL3S3IoIbXIOjhGgbdxvkbns312SigvZFUmKefXFPyjXeOTPOnDln7j7fD4Q7c55z7nk45HfPnPmfmb8jQgDyWdF0AwCaQfiBpAg/kBThB5Ii/EBShB9IivADSRF+ICnCDyT1J3XubO3atTE5OVnnLoFUZmdndfr0afez7lDht327pEclrZT0rxHxcNn6k5OTarfbw+wSQIlWq9X3ugO/7Le9UtK/SPqypGslbbd97aC/D0C9hrnm3yzpeES8FRF/kPQzSVuraQvAqA0T/vWSfrfo+Vyx7CNsT9lu2253Op0hdgegSsOEf6k3FT72+eCImImIVkS0JiYmhtgdgCoNE/45SRsWPf+cpBPDtQOgLsOE/xVJ19j+vO1PS/q6pIPVtAVg1AYe6ouIc7bvl/TvWhjq2x0Rv6qsMwAjNdQ4f0Q8I+mZinoBUCNu7wWSIvxAUoQfSIrwA0kRfiApwg8kRfiBpAg/kBThB5Ii/EBShB9IivADSRF+ICnCDyRF+IGkCD+QFOEHkiL8QFKEH0iK8ANJEX4gKcIPJEX4gaQIP5AU4QeSIvxAUoQfSIrwA0kRfiCpoWbptT0r6QNJH0o6FxGtKpoCMHpDhb/wtxFxuoLfA6BGvOwHkho2/CHpl7ZftT1VRUMA6jHsy/6bI+KE7SslPWf7zYh4cfEKxR+FKUnauHHjkLsDUJWhzvwRcaL4eUrSfkmbl1hnJiJaEdGamJgYZncAKjRw+G2vsr36wmNJWyS9UVVjAEZrmJf9V0nab/vC7/lpRPxbJV0BGLmBwx8Rb0n6iwp7AVAjhvqApAg/kBThB5Ii/EBShB9IivADSVXxqT4sYxFRWp+fny+tP/nkk6X1p556qmvt+PHjpdsePny4tM7t4sPhzA8kRfiBpAg/kBThB5Ii/EBShB9IivADSTHO///A3Nxc19qBAwdKt927d29p/dChQwP11I9Vq1aV1i+99NKR7Ruc+YG0CD+QFOEHkiL8QFKEH0iK8ANJEX4gKcb5x8DRo0dL6w899FBpff/+/V1rZ8+eLd12cnKytD49PV1aP3fuXGl9165dXWtbtmwp3Xbt2rWldQyHMz+QFOEHkiL8QFKEH0iK8ANJEX4gKcIPJNVznN/2bklfkXQqIq4rll0haa+kSUmzku6OiPdG1+Z4e+GFF0rr9957b2n97bffLq2fOXOmtD41NdW1ds8995Rue+ONN5bWe32m/siRI6X1snH+66+/vnRbjFY/Z/4fS7r9omUPSHo+Iq6R9HzxHMAy0jP8EfGipHcvWrxV0p7i8R5Jd1XcF4ARG/Sa/6qImJek4ueV1bUEoA4jf8PP9pTttu12p9MZ9e4A9GnQ8J+0vU6Sip+nuq0YETMR0YqI1sTExIC7A1C1QcN/UNKO4vEOSU9X0w6AuvQMv+0nJP2npD+zPWf7W5IelnSb7d9Iuq14DmAZ6TnOHxHbu5S+VHEvy9bp06dL65s2bSqt9/r++m3btpXW77zzzq61FSvG9z6uSy65pOkWUhvf/xkARorwA0kRfiApwg8kRfiBpAg/kJQjoradtVqtaLfbte0Po3fHHXeU1p999tmutffeK/8U+OWXXz5QT5m1Wi212233sy5nfiApwg8kRfiBpAg/kBThB5Ii/EBShB9Iiim6MZT5+fmmW8CAOPMDSRF+ICnCDyRF+IGkCD+QFOEHkiL8QFKM82Okbrrppq611atX19gJLsaZH0iK8ANJEX4gKcIPJEX4gaQIP5AU4QeS6jnOb3u3pK9IOhUR1xXLHpT0bUmdYrWdEfHMqJpEc+bm5krrx44dK62XTR++cuXKgXpCNfo58/9Y0u1LLP9hRGwq/hF8YJnpGf6IeFHSuzX0AqBGw1zz32/7qO3dttdU1hGAWgwa/l2SviBpk6R5SY90W9H2lO227Xan0+m2GoCaDRT+iDgZER9GxHlJP5K0uWTdmYhoRURrYmJi0D4BVGyg8Ntet+jpVyW9UU07AOrSz1DfE5JulbTW9pyk70u61fYmSSFpVtJ9I+wRwAj0DH9EbF9i8WMj6AVj6MCBA6X1s2fPltanp6erbAcV4g4/ICnCDyRF+IGkCD+QFOEHkiL8QFJ8dTdKvfTSS6X1FSvKzx9XX311le2gQpz5gaQIP5AU4QeSIvxAUoQfSIrwA0kRfiApxvlRan5+vrR+ww03lNY3btxYZTuoEGd+ICnCDyRF+IGkCD+QFOEHkiL8QFKEH0iK8ANJEX4gKcIPJEX4gaQIP5AU4QeSIvxAUoQfSKrn5/ltb5D0uKTPSjovaSYiHrV9haS9kiYlzUq6OyLeG12rGIX333+/tP7yyy+X1m+55ZYq20GN+jnzn5P0vYj4c0l/Jek7tq+V9ICk5yPiGknPF88BLBM9wx8R8xHxWvH4A0nHJK2XtFXSnmK1PZLuGlWTAKr3ia75bU9K+qKkw5Kuioh5aeEPhKQrq24OwOj0HX7bn5G0T9J3I6L8QvGj203ZbttudzqdQXoEMAJ9hd/2p7QQ/J9ExC+KxSdtryvq6ySdWmrbiJiJiFZEtCYmJqroGUAFeobftiU9JulYRPxgUemgpB3F4x2Snq6+PQCj0s9Xd98s6RuSXrd9pFi2U9LDkn5u+1uSfivpa6NpEaN08ODB0vqZM2dK69PT01W2gxr1DH9EHJLkLuUvVdsOgLpwhx+QFOEHkiL8QFKEH0iK8ANJEX4gKaboTm7fvn1Dbb9hw4aKOkHdOPMDSRF+ICnCDyRF+IGkCD+QFOEHkiL8QFKM86PUZZddVlrn25mWL878QFKEH0iK8ANJEX4gKcIPJEX4gaQIP5AU4/zJvfnmm6X1NWvWlNbXr19fZTuoEWd+ICnCDyRF+IGkCD+QFOEHkiL8QFKEH0iq5zi/7Q2SHpf0WUnnJc1ExKO2H5T0bUmdYtWdEfHMqBrFYB555JHSeq9x/vvuu6/KdjBG+rnJ55yk70XEa7ZXS3rV9nNF7YcR8U+jaw/AqPQMf0TMS5ovHn9g+5gkbusClrlPdM1ve1LSFyUdLhbdb/uo7d22l7wP1PaU7bbtdqfTWWoVAA3oO/y2PyNpn6TvRsT7knZJ+oKkTVp4ZbDkxWVEzEREKyJafN8bMD76Cr/tT2kh+D+JiF9IUkScjIgPI+K8pB9J2jy6NgFUrWf4bVvSY5KORcQPFi1ft2i1r0p6o/r2AIxKP+/23yzpG5Jet32kWLZT0nbbmySFpFlJjAmNoXfeeWeo7bdt21ZRJxg3/bzbf0iSlygxpg8sY9zhByRF+IGkCD+QFOEHkiL8QFKEH0jKEVHbzlqtVrTb7dr2B2TTarXUbreXGpr/GM78QFKEH0iK8ANJEX4gKcIPJEX4gaQIP5BUreP8tjuS/mfRorWSTtfWwCczrr2Na18SvQ2qyt6ujoi+vi+v1vB/bOd2OyJajTVQYlx7G9e+JHobVFO98bIfSIrwA0k1Hf6ZhvdfZlx7G9e+JHobVCO9NXrND6A5TZ/5ATSkkfDbvt32f9k+bvuBJnroxvas7ddtH7Hd6OePi2nQTtl+Y9GyK2w/Z/s3xc8lp0lrqLcHbf9vceyO2L6jod422H7B9jHbv7L9d8XyRo9dSV+NHLfaX/bbXinpvyXdJmlO0iuStkfEr2ttpAvbs5JaEdH4mLDtv5H0e0mPR8R1xbJ/lPRuRDxc/OFcExF/Pya9PSjp903P3FxMKLNu8czSku6S9E01eOxK+rpbDRy3Js78myUdj4i3IuIPkn4maWsDfYy9iHhR0rsXLd4qaU/xeI8W/vPUrktvYyEi5iPiteLxB5IuzCzd6LEr6asRTYR/vaTfLXo+p/Ga8jsk/dL2q7anmm5mCVcV06ZfmD79yob7uVjPmZvrdNHM0mNz7AaZ8bpqTYR/qa8YGqchh5sj4i8lfVnSd4qXt+hPXzM312WJmaXHwqAzXletifDPSdqw6PnnJJ1ooI8lRcSJ4ucpSfs1frMPn7wwSWrx81TD/fzROM3cvNTM0hqDYzdOM143Ef5XJF1j+/O2Py3p65IONtDHx9heVbwRI9urJG3R+M0+fFDSjuLxDklPN9jLR4zLzM3dZpZWw8du3Ga8buQmn2Io458lrZS0OyL+ofYmlmD7T7VwtpcWJjH9aZO92X5C0q1a+NTXSUnfl3RA0s8lbZT0W0lfi4ja33jr0tutWnjp+seZmy9cY9fc219L+g9Jr0s6XyzeqYXr68aOXUlf29XAceMOPyAp7vADkiL8QFKEH0iK8ANJEX4gKcIPJEX4gaQIP5DU/wHQtqOLi4bOuQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "digit = train_images[42] \n",
    " \n",
    "import matplotlib.pyplot as plt \n",
    "%matplotlib inline\n",
    "plt.imshow(digit, cmap=plt.cm.binary)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:35:40.133411Z",
     "start_time": "2019-01-11T12:35:39.789382Z"
    }
   },
   "outputs": [],
   "source": [
    "train_images = train_images.reshape((60000, 28 * 28))\n",
    "train_images = train_images.astype('float32') / 255\n",
    "\n",
    "test_images = test_images.reshape((10000, 28 * 28))\n",
    "test_images = test_images.astype('float32') / 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:35:40.474656Z",
     "start_time": "2019-01-11T12:35:40.469652Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 784)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_images.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:35:42.031054Z",
     "start_time": "2019-01-11T12:35:42.022009Z"
    }
   },
   "outputs": [],
   "source": [
    "from keras.utils import to_categorical \n",
    " \n",
    "train_labels = to_categorical(train_labels)\n",
    "test_labels = to_categorical(test_labels) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:36:02.389814Z",
     "start_time": "2019-01-11T12:35:44.043873Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "60000/60000 [==============================] - 7s 114us/step - loss: 0.2567 - acc: 0.9260\n",
      "Epoch 2/5\n",
      "60000/60000 [==============================] - 3s 44us/step - loss: 0.1019 - acc: 0.9699\n",
      "Epoch 3/5\n",
      "60000/60000 [==============================] - 3s 45us/step - loss: 0.0693 - acc: 0.9793\n",
      "Epoch 4/5\n",
      "60000/60000 [==============================] - 3s 45us/step - loss: 0.0499 - acc: 0.9847\n",
      "Epoch 5/5\n",
      "60000/60000 [==============================] - 3s 54us/step - loss: 0.0369 - acc: 0.9883\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x19a1d3c1c18>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " net.fit(train_images, train_labels, epochs=5, batch_size=128)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T12:36:02.984814Z",
     "start_time": "2019-01-11T12:36:02.402825Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000/10000 [==============================] - 1s 57us/step\n"
     ]
    }
   ],
   "source": [
    " test_loss, test_acc = net.evaluate(test_images, test_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-11T06:23:17.861161Z",
     "start_time": "2019-01-11T06:23:17.856133Z"
    }
   },
   "source": [
    "神经网络的核心组件是层（layer），它是一种数据处理模块，你可以将它看成数据过滤器。 进去一些数据，出来的数据变得更加有用。具体来说，层从输入数据中提取**表示**——我们期望这种表示有助于解决手头的问题。大多数深度学习都是简单的层链接起来的，从而实现渐进式的**数据蒸馏**（data distillation）。深度学习模型就像是数据处理的筛子，包含一系列越来越精细的 数据过滤器（即层）。 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-30T13:20:35.947370Z",
     "start_time": "2019-01-30T13:20:35.918402Z"
    }
   },
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append('D:/API/cocoapi/PythonAPI')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-30T13:20:38.598503Z",
     "start_time": "2019-01-30T13:20:38.594443Z"
    }
   },
   "outputs": [],
   "source": [
    "from simpledet.utils import generate_roidb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-30T13:20:48.367567Z",
     "start_time": "2019-01-30T13:20:48.273445Z"
    }
   },
   "outputs": [],
   "source": [
    "generate_roidb.COCO?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  },
  "nbTranslate": {
   "displayLangs": [
    "*"
   ],
   "hotkey": "alt-t",
   "langInMainMenu": true,
   "sourceLang": "en",
   "targetLang": "fr",
   "useGoogleTranslate": true
  },
  "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
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}