├── Fashion-MNIST
    └── conv_net.ipynb
└── README.md


/Fashion-MNIST/conv_net.ipynb:
--------------------------------------------------------------------------------
   1 | {
   2 |  "cells": [
   3 |   {
   4 |    "cell_type": "markdown",
   5 |    "metadata": {},
   6 |    "source": [
   7 |     "Suppress all warnings"
   8 |    ]
   9 |   },
  10 |   {
  11 |    "cell_type": "code",
  12 |    "execution_count": 357,
  13 |    "metadata": {},
  14 |    "outputs": [],
  15 |    "source": [
  16 |     "import warnings\n",
  17 |     "warnings.filterwarnings(\"ignore\")"
  18 |    ]
  19 |   },
  20 |   {
  21 |    "cell_type": "markdown",
  22 |    "metadata": {},
  23 |    "source": [
  24 |     "Import basic libraries"
  25 |    ]
  26 |   },
  27 |   {
  28 |    "cell_type": "code",
  29 |    "execution_count": 358,
  30 |    "metadata": {},
  31 |    "outputs": [],
  32 |    "source": [
  33 |     "import numpy as np\n",
  34 |     "import pandas as pd\n",
  35 |     "\n",
  36 |     "import os\n",
  37 |     "\n",
  38 |     "import matplotlib.pyplot as plt\n",
  39 |     "import matplotlib\n",
  40 |     "%matplotlib inline\n",
  41 |     "\n",
  42 |     "import seaborn as sns"
  43 |    ]
  44 |   },
  45 |   {
  46 |    "cell_type": "code",
  47 |    "execution_count": 359,
  48 |    "metadata": {},
  49 |    "outputs": [],
  50 |    "source": [
  51 |     "# To output/print multiple things from a single cell\n",
  52 |     "\n",
  53 |     "from IPython.core.interactiveshell import InteractiveShell\n",
  54 |     "InteractiveShell.ast_node_interactivity = \"all\""
  55 |    ]
  56 |   },
  57 |   {
  58 |    "cell_type": "code",
  59 |    "execution_count": 360,
  60 |    "metadata": {},
  61 |    "outputs": [],
  62 |    "source": [
  63 |     "from sklearn.model_selection import train_test_split, cross_val_predict\n",
  64 |     "from sklearn.metrics import accuracy_score"
  65 |    ]
  66 |   },
  67 |   {
  68 |    "cell_type": "code",
  69 |    "execution_count": 361,
  70 |    "metadata": {},
  71 |    "outputs": [],
  72 |    "source": [
  73 |     "# specify which certain GPUs are in use (before import tensorflow)\n",
  74 |     "os.environ['CUDA_DEVICE_ORDER'] = 'PCI_BUS_ID'\n",
  75 |     "os.environ['CUDA_VISIBLE_DEVICES'] = '0'\n",
  76 |     "os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'"
  77 |    ]
  78 |   },
  79 |   {
  80 |    "cell_type": "code",
  81 |    "execution_count": 362,
  82 |    "metadata": {},
  83 |    "outputs": [],
  84 |    "source": [
  85 |     "import tensorflow as tf\n",
  86 |     "from tensorflow.keras.preprocessing.image import ImageDataGenerator"
  87 |    ]
  88 |   },
  89 |   {
  90 |    "cell_type": "markdown",
  91 |    "metadata": {},
  92 |    "source": [
  93 |     "### Load Train/Test Data"
  94 |    ]
  95 |   },
  96 |   {
  97 |    "cell_type": "code",
  98 |    "execution_count": 363,
  99 |    "metadata": {},
 100 |    "outputs": [],
 101 |    "source": [
 102 |     "train = pd.read_csv('fashion-mnist_train.csv')\n",
 103 |     "test = pd.read_csv('fashion-mnist_test.csv')"
 104 |    ]
 105 |   },
 106 |   {
 107 |    "cell_type": "code",
 108 |    "execution_count": 364,
 109 |    "metadata": {},
 110 |    "outputs": [
 111 |     {
 112 |      "data": {
 113 |       "text/plain": [
 114 |        "(60000, 785)"
 115 |       ]
 116 |      },
 117 |      "execution_count": 364,
 118 |      "metadata": {},
 119 |      "output_type": "execute_result"
 120 |     },
 121 |     {
 122 |      "data": {
 123 |       "text/plain": [
 124 |        "(10000, 785)"
 125 |       ]
 126 |      },
 127 |      "execution_count": 364,
 128 |      "metadata": {},
 129 |      "output_type": "execute_result"
 130 |     }
 131 |    ],
 132 |    "source": [
 133 |     "train.shape\n",
 134 |     "test.shape"
 135 |    ]
 136 |   },
 137 |   {
 138 |    "cell_type": "code",
 139 |    "execution_count": 365,
 140 |    "metadata": {},
 141 |    "outputs": [
 142 |     {
 143 |      "data": {
 144 |       "text/html": [
 145 |        "<div>\n",
 146 |        "<style scoped>\n",
 147 |        "    .dataframe tbody tr th:only-of-type {\n",
 148 |        "        vertical-align: middle;\n",
 149 |        "    }\n",
 150 |        "\n",
 151 |        "    .dataframe tbody tr th {\n",
 152 |        "        vertical-align: top;\n",
 153 |        "    }\n",
 154 |        "\n",
 155 |        "    .dataframe thead th {\n",
 156 |        "        text-align: right;\n",
 157 |        "    }\n",
 158 |        "</style>\n",
 159 |        "<table border=\"1\" class=\"dataframe\">\n",
 160 |        "  <thead>\n",
 161 |        "    <tr style=\"text-align: right;\">\n",
 162 |        "      <th></th>\n",
 163 |        "      <th>label</th>\n",
 164 |        "      <th>pixel1</th>\n",
 165 |        "      <th>pixel2</th>\n",
 166 |        "      <th>pixel3</th>\n",
 167 |        "      <th>pixel4</th>\n",
 168 |        "      <th>pixel5</th>\n",
 169 |        "      <th>pixel6</th>\n",
 170 |        "      <th>pixel7</th>\n",
 171 |        "      <th>pixel8</th>\n",
 172 |        "      <th>pixel9</th>\n",
 173 |        "      <th>...</th>\n",
 174 |        "      <th>pixel775</th>\n",
 175 |        "      <th>pixel776</th>\n",
 176 |        "      <th>pixel777</th>\n",
 177 |        "      <th>pixel778</th>\n",
 178 |        "      <th>pixel779</th>\n",
 179 |        "      <th>pixel780</th>\n",
 180 |        "      <th>pixel781</th>\n",
 181 |        "      <th>pixel782</th>\n",
 182 |        "      <th>pixel783</th>\n",
 183 |        "      <th>pixel784</th>\n",
 184 |        "    </tr>\n",
 185 |        "  </thead>\n",
 186 |        "  <tbody>\n",
 187 |        "    <tr>\n",
 188 |        "      <th>0</th>\n",
 189 |        "      <td>2</td>\n",
 190 |        "      <td>0</td>\n",
 191 |        "      <td>0</td>\n",
 192 |        "      <td>0</td>\n",
 193 |        "      <td>0</td>\n",
 194 |        "      <td>0</td>\n",
 195 |        "      <td>0</td>\n",
 196 |        "      <td>0</td>\n",
 197 |        "      <td>0</td>\n",
 198 |        "      <td>0</td>\n",
 199 |        "      <td>...</td>\n",
 200 |        "      <td>0</td>\n",
 201 |        "      <td>0</td>\n",
 202 |        "      <td>0</td>\n",
 203 |        "      <td>0</td>\n",
 204 |        "      <td>0</td>\n",
 205 |        "      <td>0</td>\n",
 206 |        "      <td>0</td>\n",
 207 |        "      <td>0</td>\n",
 208 |        "      <td>0</td>\n",
 209 |        "      <td>0</td>\n",
 210 |        "    </tr>\n",
 211 |        "    <tr>\n",
 212 |        "      <th>1</th>\n",
 213 |        "      <td>9</td>\n",
 214 |        "      <td>0</td>\n",
 215 |        "      <td>0</td>\n",
 216 |        "      <td>0</td>\n",
 217 |        "      <td>0</td>\n",
 218 |        "      <td>0</td>\n",
 219 |        "      <td>0</td>\n",
 220 |        "      <td>0</td>\n",
 221 |        "      <td>0</td>\n",
 222 |        "      <td>0</td>\n",
 223 |        "      <td>...</td>\n",
 224 |        "      <td>0</td>\n",
 225 |        "      <td>0</td>\n",
 226 |        "      <td>0</td>\n",
 227 |        "      <td>0</td>\n",
 228 |        "      <td>0</td>\n",
 229 |        "      <td>0</td>\n",
 230 |        "      <td>0</td>\n",
 231 |        "      <td>0</td>\n",
 232 |        "      <td>0</td>\n",
 233 |        "      <td>0</td>\n",
 234 |        "    </tr>\n",
 235 |        "    <tr>\n",
 236 |        "      <th>2</th>\n",
 237 |        "      <td>6</td>\n",
 238 |        "      <td>0</td>\n",
 239 |        "      <td>0</td>\n",
 240 |        "      <td>0</td>\n",
 241 |        "      <td>0</td>\n",
 242 |        "      <td>0</td>\n",
 243 |        "      <td>0</td>\n",
 244 |        "      <td>0</td>\n",
 245 |        "      <td>5</td>\n",
 246 |        "      <td>0</td>\n",
 247 |        "      <td>...</td>\n",
 248 |        "      <td>0</td>\n",
 249 |        "      <td>0</td>\n",
 250 |        "      <td>0</td>\n",
 251 |        "      <td>30</td>\n",
 252 |        "      <td>43</td>\n",
 253 |        "      <td>0</td>\n",
 254 |        "      <td>0</td>\n",
 255 |        "      <td>0</td>\n",
 256 |        "      <td>0</td>\n",
 257 |        "      <td>0</td>\n",
 258 |        "    </tr>\n",
 259 |        "    <tr>\n",
 260 |        "      <th>3</th>\n",
 261 |        "      <td>0</td>\n",
 262 |        "      <td>0</td>\n",
 263 |        "      <td>0</td>\n",
 264 |        "      <td>0</td>\n",
 265 |        "      <td>1</td>\n",
 266 |        "      <td>2</td>\n",
 267 |        "      <td>0</td>\n",
 268 |        "      <td>0</td>\n",
 269 |        "      <td>0</td>\n",
 270 |        "      <td>0</td>\n",
 271 |        "      <td>...</td>\n",
 272 |        "      <td>3</td>\n",
 273 |        "      <td>0</td>\n",
 274 |        "      <td>0</td>\n",
 275 |        "      <td>0</td>\n",
 276 |        "      <td>0</td>\n",
 277 |        "      <td>1</td>\n",
 278 |        "      <td>0</td>\n",
 279 |        "      <td>0</td>\n",
 280 |        "      <td>0</td>\n",
 281 |        "      <td>0</td>\n",
 282 |        "    </tr>\n",
 283 |        "    <tr>\n",
 284 |        "      <th>4</th>\n",
 285 |        "      <td>3</td>\n",
 286 |        "      <td>0</td>\n",
 287 |        "      <td>0</td>\n",
 288 |        "      <td>0</td>\n",
 289 |        "      <td>0</td>\n",
 290 |        "      <td>0</td>\n",
 291 |        "      <td>0</td>\n",
 292 |        "      <td>0</td>\n",
 293 |        "      <td>0</td>\n",
 294 |        "      <td>0</td>\n",
 295 |        "      <td>...</td>\n",
 296 |        "      <td>0</td>\n",
 297 |        "      <td>0</td>\n",
 298 |        "      <td>0</td>\n",
 299 |        "      <td>0</td>\n",
 300 |        "      <td>0</td>\n",
 301 |        "      <td>0</td>\n",
 302 |        "      <td>0</td>\n",
 303 |        "      <td>0</td>\n",
 304 |        "      <td>0</td>\n",
 305 |        "      <td>0</td>\n",
 306 |        "    </tr>\n",
 307 |        "  </tbody>\n",
 308 |        "</table>\n",
 309 |        "<p>5 rows × 785 columns</p>\n",
 310 |        "</div>"
 311 |       ],
 312 |       "text/plain": [
 313 |        "   label  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  pixel8  \\\n",
 314 |        "0      2       0       0       0       0       0       0       0       0   \n",
 315 |        "1      9       0       0       0       0       0       0       0       0   \n",
 316 |        "2      6       0       0       0       0       0       0       0       5   \n",
 317 |        "3      0       0       0       0       1       2       0       0       0   \n",
 318 |        "4      3       0       0       0       0       0       0       0       0   \n",
 319 |        "\n",
 320 |        "   pixel9  ...  pixel775  pixel776  pixel777  pixel778  pixel779  pixel780  \\\n",
 321 |        "0       0  ...         0         0         0         0         0         0   \n",
 322 |        "1       0  ...         0         0         0         0         0         0   \n",
 323 |        "2       0  ...         0         0         0        30        43         0   \n",
 324 |        "3       0  ...         3         0         0         0         0         1   \n",
 325 |        "4       0  ...         0         0         0         0         0         0   \n",
 326 |        "\n",
 327 |        "   pixel781  pixel782  pixel783  pixel784  \n",
 328 |        "0         0         0         0         0  \n",
 329 |        "1         0         0         0         0  \n",
 330 |        "2         0         0         0         0  \n",
 331 |        "3         0         0         0         0  \n",
 332 |        "4         0         0         0         0  \n",
 333 |        "\n",
 334 |        "[5 rows x 785 columns]"
 335 |       ]
 336 |      },
 337 |      "execution_count": 365,
 338 |      "metadata": {},
 339 |      "output_type": "execute_result"
 340 |     },
 341 |     {
 342 |      "data": {
 343 |       "text/html": [
 344 |        "<div>\n",
 345 |        "<style scoped>\n",
 346 |        "    .dataframe tbody tr th:only-of-type {\n",
 347 |        "        vertical-align: middle;\n",
 348 |        "    }\n",
 349 |        "\n",
 350 |        "    .dataframe tbody tr th {\n",
 351 |        "        vertical-align: top;\n",
 352 |        "    }\n",
 353 |        "\n",
 354 |        "    .dataframe thead th {\n",
 355 |        "        text-align: right;\n",
 356 |        "    }\n",
 357 |        "</style>\n",
 358 |        "<table border=\"1\" class=\"dataframe\">\n",
 359 |        "  <thead>\n",
 360 |        "    <tr style=\"text-align: right;\">\n",
 361 |        "      <th></th>\n",
 362 |        "      <th>label</th>\n",
 363 |        "      <th>pixel1</th>\n",
 364 |        "      <th>pixel2</th>\n",
 365 |        "      <th>pixel3</th>\n",
 366 |        "      <th>pixel4</th>\n",
 367 |        "      <th>pixel5</th>\n",
 368 |        "      <th>pixel6</th>\n",
 369 |        "      <th>pixel7</th>\n",
 370 |        "      <th>pixel8</th>\n",
 371 |        "      <th>pixel9</th>\n",
 372 |        "      <th>...</th>\n",
 373 |        "      <th>pixel775</th>\n",
 374 |        "      <th>pixel776</th>\n",
 375 |        "      <th>pixel777</th>\n",
 376 |        "      <th>pixel778</th>\n",
 377 |        "      <th>pixel779</th>\n",
 378 |        "      <th>pixel780</th>\n",
 379 |        "      <th>pixel781</th>\n",
 380 |        "      <th>pixel782</th>\n",
 381 |        "      <th>pixel783</th>\n",
 382 |        "      <th>pixel784</th>\n",
 383 |        "    </tr>\n",
 384 |        "  </thead>\n",
 385 |        "  <tbody>\n",
 386 |        "    <tr>\n",
 387 |        "      <th>0</th>\n",
 388 |        "      <td>0</td>\n",
 389 |        "      <td>0</td>\n",
 390 |        "      <td>0</td>\n",
 391 |        "      <td>0</td>\n",
 392 |        "      <td>0</td>\n",
 393 |        "      <td>0</td>\n",
 394 |        "      <td>0</td>\n",
 395 |        "      <td>0</td>\n",
 396 |        "      <td>9</td>\n",
 397 |        "      <td>8</td>\n",
 398 |        "      <td>...</td>\n",
 399 |        "      <td>103</td>\n",
 400 |        "      <td>87</td>\n",
 401 |        "      <td>56</td>\n",
 402 |        "      <td>0</td>\n",
 403 |        "      <td>0</td>\n",
 404 |        "      <td>0</td>\n",
 405 |        "      <td>0</td>\n",
 406 |        "      <td>0</td>\n",
 407 |        "      <td>0</td>\n",
 408 |        "      <td>0</td>\n",
 409 |        "    </tr>\n",
 410 |        "    <tr>\n",
 411 |        "      <th>1</th>\n",
 412 |        "      <td>1</td>\n",
 413 |        "      <td>0</td>\n",
 414 |        "      <td>0</td>\n",
 415 |        "      <td>0</td>\n",
 416 |        "      <td>0</td>\n",
 417 |        "      <td>0</td>\n",
 418 |        "      <td>0</td>\n",
 419 |        "      <td>0</td>\n",
 420 |        "      <td>0</td>\n",
 421 |        "      <td>0</td>\n",
 422 |        "      <td>...</td>\n",
 423 |        "      <td>34</td>\n",
 424 |        "      <td>0</td>\n",
 425 |        "      <td>0</td>\n",
 426 |        "      <td>0</td>\n",
 427 |        "      <td>0</td>\n",
 428 |        "      <td>0</td>\n",
 429 |        "      <td>0</td>\n",
 430 |        "      <td>0</td>\n",
 431 |        "      <td>0</td>\n",
 432 |        "      <td>0</td>\n",
 433 |        "    </tr>\n",
 434 |        "    <tr>\n",
 435 |        "      <th>2</th>\n",
 436 |        "      <td>2</td>\n",
 437 |        "      <td>0</td>\n",
 438 |        "      <td>0</td>\n",
 439 |        "      <td>0</td>\n",
 440 |        "      <td>0</td>\n",
 441 |        "      <td>0</td>\n",
 442 |        "      <td>0</td>\n",
 443 |        "      <td>14</td>\n",
 444 |        "      <td>53</td>\n",
 445 |        "      <td>99</td>\n",
 446 |        "      <td>...</td>\n",
 447 |        "      <td>0</td>\n",
 448 |        "      <td>0</td>\n",
 449 |        "      <td>0</td>\n",
 450 |        "      <td>0</td>\n",
 451 |        "      <td>63</td>\n",
 452 |        "      <td>53</td>\n",
 453 |        "      <td>31</td>\n",
 454 |        "      <td>0</td>\n",
 455 |        "      <td>0</td>\n",
 456 |        "      <td>0</td>\n",
 457 |        "    </tr>\n",
 458 |        "    <tr>\n",
 459 |        "      <th>3</th>\n",
 460 |        "      <td>2</td>\n",
 461 |        "      <td>0</td>\n",
 462 |        "      <td>0</td>\n",
 463 |        "      <td>0</td>\n",
 464 |        "      <td>0</td>\n",
 465 |        "      <td>0</td>\n",
 466 |        "      <td>0</td>\n",
 467 |        "      <td>0</td>\n",
 468 |        "      <td>0</td>\n",
 469 |        "      <td>0</td>\n",
 470 |        "      <td>...</td>\n",
 471 |        "      <td>137</td>\n",
 472 |        "      <td>126</td>\n",
 473 |        "      <td>140</td>\n",
 474 |        "      <td>0</td>\n",
 475 |        "      <td>133</td>\n",
 476 |        "      <td>224</td>\n",
 477 |        "      <td>222</td>\n",
 478 |        "      <td>56</td>\n",
 479 |        "      <td>0</td>\n",
 480 |        "      <td>0</td>\n",
 481 |        "    </tr>\n",
 482 |        "    <tr>\n",
 483 |        "      <th>4</th>\n",
 484 |        "      <td>3</td>\n",
 485 |        "      <td>0</td>\n",
 486 |        "      <td>0</td>\n",
 487 |        "      <td>0</td>\n",
 488 |        "      <td>0</td>\n",
 489 |        "      <td>0</td>\n",
 490 |        "      <td>0</td>\n",
 491 |        "      <td>0</td>\n",
 492 |        "      <td>0</td>\n",
 493 |        "      <td>0</td>\n",
 494 |        "      <td>...</td>\n",
 495 |        "      <td>0</td>\n",
 496 |        "      <td>0</td>\n",
 497 |        "      <td>0</td>\n",
 498 |        "      <td>0</td>\n",
 499 |        "      <td>0</td>\n",
 500 |        "      <td>0</td>\n",
 501 |        "      <td>0</td>\n",
 502 |        "      <td>0</td>\n",
 503 |        "      <td>0</td>\n",
 504 |        "      <td>0</td>\n",
 505 |        "    </tr>\n",
 506 |        "  </tbody>\n",
 507 |        "</table>\n",
 508 |        "<p>5 rows × 785 columns</p>\n",
 509 |        "</div>"
 510 |       ],
 511 |       "text/plain": [
 512 |        "   label  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7  pixel8  \\\n",
 513 |        "0      0       0       0       0       0       0       0       0       9   \n",
 514 |        "1      1       0       0       0       0       0       0       0       0   \n",
 515 |        "2      2       0       0       0       0       0       0      14      53   \n",
 516 |        "3      2       0       0       0       0       0       0       0       0   \n",
 517 |        "4      3       0       0       0       0       0       0       0       0   \n",
 518 |        "\n",
 519 |        "   pixel9  ...  pixel775  pixel776  pixel777  pixel778  pixel779  pixel780  \\\n",
 520 |        "0       8  ...       103        87        56         0         0         0   \n",
 521 |        "1       0  ...        34         0         0         0         0         0   \n",
 522 |        "2      99  ...         0         0         0         0        63        53   \n",
 523 |        "3       0  ...       137       126       140         0       133       224   \n",
 524 |        "4       0  ...         0         0         0         0         0         0   \n",
 525 |        "\n",
 526 |        "   pixel781  pixel782  pixel783  pixel784  \n",
 527 |        "0         0         0         0         0  \n",
 528 |        "1         0         0         0         0  \n",
 529 |        "2        31         0         0         0  \n",
 530 |        "3       222        56         0         0  \n",
 531 |        "4         0         0         0         0  \n",
 532 |        "\n",
 533 |        "[5 rows x 785 columns]"
 534 |       ]
 535 |      },
 536 |      "execution_count": 365,
 537 |      "metadata": {},
 538 |      "output_type": "execute_result"
 539 |     }
 540 |    ],
 541 |    "source": [
 542 |     "train.head()\n",
 543 |     "test.head()"
 544 |    ]
 545 |   },
 546 |   {
 547 |    "cell_type": "code",
 548 |    "execution_count": 366,
 549 |    "metadata": {},
 550 |    "outputs": [],
 551 |    "source": [
 552 |     "# Separate out features and target\n",
 553 |     "\n",
 554 |     "X_train = train.copy()\n",
 555 |     "y_train = train['label']\n",
 556 |     "X_train.drop('label', axis=1, inplace=True)\n",
 557 |     "\n",
 558 |     "X_test = test.copy()\n",
 559 |     "y_test = test['label']\n",
 560 |     "X_test.drop('label', axis=1, inplace=True)"
 561 |    ]
 562 |   },
 563 |   {
 564 |    "cell_type": "markdown",
 565 |    "metadata": {},
 566 |    "source": [
 567 |     "<b> Plot class counts for target</b>"
 568 |    ]
 569 |   },
 570 |   {
 571 |    "cell_type": "code",
 572 |    "execution_count": 367,
 573 |    "metadata": {},
 574 |    "outputs": [
 575 |     {
 576 |      "data": {
 577 |       "image/png": "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\n",
 578 |       "text/plain": [
 579 |        "<Figure size 432x288 with 1 Axes>"
 580 |       ]
 581 |      },
 582 |      "metadata": {},
 583 |      "output_type": "display_data"
 584 |     }
 585 |    ],
 586 |    "source": [
 587 |     "_ = sns.countplot(y_train)"
 588 |    ]
 589 |   },
 590 |   {
 591 |    "cell_type": "code",
 592 |    "execution_count": 368,
 593 |    "metadata": {
 594 |     "scrolled": true
 595 |    },
 596 |    "outputs": [
 597 |     {
 598 |      "data": {
 599 |       "image/png": "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\n",
 600 |       "text/plain": [
 601 |        "<Figure size 432x288 with 1 Axes>"
 602 |       ]
 603 |      },
 604 |      "metadata": {},
 605 |      "output_type": "display_data"
 606 |     }
 607 |    ],
 608 |    "source": [
 609 |     "_ = sns.countplot(y_test)"
 610 |    ]
 611 |   },
 612 |   {
 613 |    "cell_type": "markdown",
 614 |    "metadata": {},
 615 |    "source": [
 616 |     "<b>Plot a target image</b>"
 617 |    ]
 618 |   },
 619 |   {
 620 |    "cell_type": "code",
 621 |    "execution_count": 369,
 622 |    "metadata": {},
 623 |    "outputs": [],
 624 |    "source": [
 625 |     "def plot_item(data):\n",
 626 |     "    image = data.reshape(28, 28)\n",
 627 |     "    plt.imshow(image, cmap=matplotlib.cm.binary,\n",
 628 |     "               interpolation=\"nearest\")\n",
 629 |     "    plt.axis(\"off\")"
 630 |    ]
 631 |   },
 632 |   {
 633 |    "cell_type": "code",
 634 |    "execution_count": 370,
 635 |    "metadata": {},
 636 |    "outputs": [
 637 |     {
 638 |      "data": {
 639 |       "image/png": "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\n",
 640 |       "text/plain": [
 641 |        "<Figure size 432x288 with 1 Axes>"
 642 |       ]
 643 |      },
 644 |      "metadata": {},
 645 |      "output_type": "display_data"
 646 |     }
 647 |    ],
 648 |    "source": [
 649 |     "plot_item(np.array(X_train.iloc[2]))"
 650 |    ]
 651 |   },
 652 |   {
 653 |    "cell_type": "markdown",
 654 |    "metadata": {},
 655 |    "source": [
 656 |     "<b>Plot multiple target images</b>"
 657 |    ]
 658 |   },
 659 |   {
 660 |    "cell_type": "code",
 661 |    "execution_count": 371,
 662 |    "metadata": {},
 663 |    "outputs": [],
 664 |    "source": [
 665 |     "def plot_items(instances, images_per_row=10, **options):\n",
 666 |     "    size = 28\n",
 667 |     "    images_per_row = min(len(instances), images_per_row)\n",
 668 |     "    images = [instance.reshape(size,size) for instance in instances]\n",
 669 |     "    n_rows = (len(instances) - 1) // images_per_row + 1\n",
 670 |     "    row_images = []\n",
 671 |     "    n_empty = n_rows * images_per_row - len(instances)\n",
 672 |     "    images.append(np.zeros((size, size * n_empty)))\n",
 673 |     "    for row in range(n_rows):\n",
 674 |     "        rimages = images[row * images_per_row : (row + 1) * images_per_row]\n",
 675 |     "        row_images.append(np.concatenate(rimages, axis=1))\n",
 676 |     "    image = np.concatenate(row_images, axis=0)\n",
 677 |     "    plt.imshow(image, cmap=matplotlib.cm.binary, **options)\n",
 678 |     "    plt.axis(\"off\")"
 679 |    ]
 680 |   },
 681 |   {
 682 |    "cell_type": "code",
 683 |    "execution_count": 372,
 684 |    "metadata": {},
 685 |    "outputs": [
 686 |     {
 687 |      "data": {
 688 |       "text/plain": [
 689 |        "<Figure size 1080x1080 with 0 Axes>"
 690 |       ]
 691 |      },
 692 |      "execution_count": 372,
 693 |      "metadata": {},
 694 |      "output_type": "execute_result"
 695 |     },
 696 |     {
 697 |      "data": {
 698 |       "image/png": "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\n",
 699 |       "text/plain": [
 700 |        "<Figure size 1080x1080 with 1 Axes>"
 701 |       ]
 702 |      },
 703 |      "metadata": {},
 704 |      "output_type": "display_data"
 705 |     }
 706 |    ],
 707 |    "source": [
 708 |     "plt.figure(figsize=(15,15))\n",
 709 |     "example_images = np.array(X_train.iloc[:100])\n",
 710 |     "plot_items(example_images, images_per_row=10)\n",
 711 |     "#save_fig(\"more_items_plot\")\n",
 712 |     "plt.show()"
 713 |    ]
 714 |   },
 715 |   {
 716 |    "cell_type": "markdown",
 717 |    "metadata": {},
 718 |    "source": [
 719 |     "<b>Normalize the gray scale pixel values between 0-1</b>"
 720 |    ]
 721 |   },
 722 |   {
 723 |    "cell_type": "code",
 724 |    "execution_count": 373,
 725 |    "metadata": {},
 726 |    "outputs": [],
 727 |    "source": [
 728 |     "X_train = np.array(X_train) / 255.0\n",
 729 |     "X_test = np.array(X_test) / 255.0"
 730 |    ]
 731 |   },
 732 |   {
 733 |    "cell_type": "markdown",
 734 |    "metadata": {},
 735 |    "source": [
 736 |     "<b>Re-shape the train/test inputs to the format (samples, height, width, channels) for the DL model</b>"
 737 |    ]
 738 |   },
 739 |   {
 740 |    "cell_type": "code",
 741 |    "execution_count": 374,
 742 |    "metadata": {},
 743 |    "outputs": [],
 744 |    "source": [
 745 |     "X_train = X_train.reshape((X_train.shape[0], 28, 28, 1))\n",
 746 |     "X_test = X_test.reshape((X_test.shape[0], 28, 28, 1))"
 747 |    ]
 748 |   },
 749 |   {
 750 |    "cell_type": "markdown",
 751 |    "metadata": {},
 752 |    "source": [
 753 |     "<b>One-hot encode categorical target values</b>"
 754 |    ]
 755 |   },
 756 |   {
 757 |    "cell_type": "code",
 758 |    "execution_count": 375,
 759 |    "metadata": {},
 760 |    "outputs": [],
 761 |    "source": [
 762 |     "from keras.utils import to_categorical\n",
 763 |     "y_train_enc = to_categorical(y_train)\n",
 764 |     "y_test_enc = to_categorical(y_test)"
 765 |    ]
 766 |   },
 767 |   {
 768 |    "cell_type": "markdown",
 769 |    "metadata": {},
 770 |    "source": [
 771 |     "## Define the model"
 772 |    ]
 773 |   },
 774 |   {
 775 |    "cell_type": "code",
 776 |    "execution_count": 376,
 777 |    "metadata": {},
 778 |    "outputs": [],
 779 |    "source": [
 780 |     "model = tf.keras.models.Sequential([\n",
 781 |     "    tf.keras.layers.Conv2D(32, (5,5), activation='relu', padding = 'Same', input_shape=(28, 28, 1)),\n",
 782 |     "    tf.keras.layers.Conv2D(32, (5,5), padding = 'Same', activation='relu'),\n",
 783 |     "    tf.keras.layers.MaxPooling2D(2,2),\n",
 784 |     "    tf.keras.layers.Dropout(0.25),\n",
 785 |     "    tf.keras.layers.Conv2D(64, (3,3), padding = 'Same', activation='relu'),\n",
 786 |     "    tf.keras.layers.Conv2D(64, (3,3), padding = 'Same', activation='relu'),\n",
 787 |     "    tf.keras.layers.MaxPooling2D(2,2),\n",
 788 |     "    tf.keras.layers.Dropout(0.25),\n",
 789 |     "    tf.keras.layers.Flatten(),\n",
 790 |     "    tf.keras.layers.Dense(256, activation='relu'), \n",
 791 |     "    tf.keras.layers.Dropout(0.4),\n",
 792 |     "    tf.keras.layers.Dense(10, activation='softmax')    \n",
 793 |     "])\n",
 794 |     "\n",
 795 |     "model.compile(optimizer='adam',\n",
 796 |     "              loss='categorical_crossentropy',\n",
 797 |     "              metrics=['accuracy'])"
 798 |    ]
 799 |   },
 800 |   {
 801 |    "cell_type": "code",
 802 |    "execution_count": 377,
 803 |    "metadata": {},
 804 |    "outputs": [
 805 |     {
 806 |      "name": "stdout",
 807 |      "output_type": "stream",
 808 |      "text": [
 809 |       "_________________________________________________________________\n",
 810 |       "Layer (type)                 Output Shape              Param #   \n",
 811 |       "=================================================================\n",
 812 |       "conv2d_50 (Conv2D)           (None, 28, 28, 32)        832       \n",
 813 |       "_________________________________________________________________\n",
 814 |       "conv2d_51 (Conv2D)           (None, 28, 28, 32)        25632     \n",
 815 |       "_________________________________________________________________\n",
 816 |       "max_pooling2d_33 (MaxPooling (None, 14, 14, 32)        0         \n",
 817 |       "_________________________________________________________________\n",
 818 |       "dropout_30 (Dropout)         (None, 14, 14, 32)        0         \n",
 819 |       "_________________________________________________________________\n",
 820 |       "conv2d_52 (Conv2D)           (None, 14, 14, 64)        18496     \n",
 821 |       "_________________________________________________________________\n",
 822 |       "conv2d_53 (Conv2D)           (None, 14, 14, 64)        36928     \n",
 823 |       "_________________________________________________________________\n",
 824 |       "max_pooling2d_34 (MaxPooling (None, 7, 7, 64)          0         \n",
 825 |       "_________________________________________________________________\n",
 826 |       "dropout_31 (Dropout)         (None, 7, 7, 64)          0         \n",
 827 |       "_________________________________________________________________\n",
 828 |       "flatten_15 (Flatten)         (None, 3136)              0         \n",
 829 |       "_________________________________________________________________\n",
 830 |       "dense_31 (Dense)             (None, 256)               803072    \n",
 831 |       "_________________________________________________________________\n",
 832 |       "dropout_32 (Dropout)         (None, 256)               0         \n",
 833 |       "_________________________________________________________________\n",
 834 |       "dense_32 (Dense)             (None, 10)                2570      \n",
 835 |       "=================================================================\n",
 836 |       "Total params: 887,530\n",
 837 |       "Trainable params: 887,530\n",
 838 |       "Non-trainable params: 0\n",
 839 |       "_________________________________________________________________\n"
 840 |      ]
 841 |     }
 842 |    ],
 843 |    "source": [
 844 |     "model.summary()"
 845 |    ]
 846 |   },
 847 |   {
 848 |    "cell_type": "code",
 849 |    "execution_count": 378,
 850 |    "metadata": {},
 851 |    "outputs": [],
 852 |    "source": [
 853 |     "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
 854 |     "train_datagen = ImageDataGenerator()\n",
 855 |     "\n",
 856 |     "val_datagen = ImageDataGenerator()\n",
 857 |     "\n",
 858 |     "train_generator = train_datagen.flow(X_train, y_train_enc, batch_size=32)\n",
 859 |     "val_generator = val_datagen.flow(X_test, y_test_enc, batch_size=32)"
 860 |    ]
 861 |   },
 862 |   {
 863 |    "cell_type": "code",
 864 |    "execution_count": 379,
 865 |    "metadata": {},
 866 |    "outputs": [],
 867 |    "source": [
 868 |     "class myCallback(tf.keras.callbacks.Callback):\n",
 869 |     "    def on_epoch_end(self, epoch, logs={}):\n",
 870 |     "        if(logs.get('val_acc') >= 0.9970):\n",
 871 |     "            print(\"\\nReached 99.70% accuracy so cancelling training!\")\n",
 872 |     "            self.model.stop_training = True\n",
 873 |     "callbacks = myCallback()"
 874 |    ]
 875 |   },
 876 |   {
 877 |    "cell_type": "code",
 878 |    "execution_count": 380,
 879 |    "metadata": {},
 880 |    "outputs": [],
 881 |    "source": [
 882 |     "# callbacks_list = [\n",
 883 |     "#     tf.keras.callbacks.EarlyStopping(monitor='acc', patience=1),\n",
 884 |     "# ]\n",
 885 |     "\n",
 886 |     "callbacks_list = [\n",
 887 |     "    tf.keras.callbacks.ModelCheckpoint(filepath='my_model.h5', monitor='val_loss', save_best_only=True)\n",
 888 |     "]"
 889 |    ]
 890 |   },
 891 |   {
 892 |    "cell_type": "code",
 893 |    "execution_count": null,
 894 |    "metadata": {},
 895 |    "outputs": [],
 896 |    "source": []
 897 |   },
 898 |   {
 899 |    "cell_type": "code",
 900 |    "execution_count": 381,
 901 |    "metadata": {
 902 |     "scrolled": true
 903 |    },
 904 |    "outputs": [
 905 |     {
 906 |      "name": "stdout",
 907 |      "output_type": "stream",
 908 |      "text": [
 909 |       "Epoch 1/10\n",
 910 |       "313/313 [==============================] - 5s 16ms/step - loss: 0.2842 - acc: 0.8946\n",
 911 |       "1875/1875 [==============================] - 63s 34ms/step - loss: 0.4594 - acc: 0.8327 - val_loss: 0.2842 - val_acc: 0.8946\n",
 912 |       "Epoch 2/10\n",
 913 |       "313/313 [==============================] - 2s 7ms/step - loss: 0.2430 - acc: 0.9060\n",
 914 |       "1875/1875 [==============================] - 42s 23ms/step - loss: 0.2994 - acc: 0.8907 - val_loss: 0.2430 - val_acc: 0.9060\n",
 915 |       "Epoch 3/10\n",
 916 |       "313/313 [==============================] - 4s 12ms/step - loss: 0.2147 - acc: 0.9205\n",
 917 |       "1875/1875 [==============================] - 44s 24ms/step - loss: 0.2608 - acc: 0.9046 - val_loss: 0.2147 - val_acc: 0.9205\n",
 918 |       "Epoch 4/10\n",
 919 |       "313/313 [==============================] - 3s 8ms/step - loss: 0.2177 - acc: 0.9176\n",
 920 |       "1875/1875 [==============================] - 44s 23ms/step - loss: 0.2379 - acc: 0.9132 - val_loss: 0.2177 - val_acc: 0.9176\n",
 921 |       "Epoch 5/10\n",
 922 |       "313/313 [==============================] - 4s 12ms/step - loss: 0.1991 - acc: 0.9250\n",
 923 |       "1875/1875 [==============================] - 45s 24ms/step - loss: 0.2230 - acc: 0.9177 - val_loss: 0.1991 - val_acc: 0.9250\n",
 924 |       "Epoch 6/10\n",
 925 |       "313/313 [==============================] - 4s 13ms/step - loss: 0.2041 - acc: 0.9231\n",
 926 |       "1875/1875 [==============================] - 52s 28ms/step - loss: 0.2090 - acc: 0.9229 - val_loss: 0.2041 - val_acc: 0.9231\n",
 927 |       "Epoch 7/10\n",
 928 |       "313/313 [==============================] - 3s 8ms/step - loss: 0.1891 - acc: 0.9290\n",
 929 |       "1875/1875 [==============================] - 42s 22ms/step - loss: 0.1999 - acc: 0.9259 - val_loss: 0.1891 - val_acc: 0.9290\n",
 930 |       "Epoch 8/10\n",
 931 |       "313/313 [==============================] - 3s 10ms/step - loss: 0.1890 - acc: 0.9296\n",
 932 |       "1875/1875 [==============================] - 45s 24ms/step - loss: 0.1900 - acc: 0.9290 - val_loss: 0.1890 - val_acc: 0.9296\n",
 933 |       "Epoch 9/10\n",
 934 |       "313/313 [==============================] - 2s 7ms/step - loss: 0.1857 - acc: 0.9301\n",
 935 |       "1875/1875 [==============================] - 43s 23ms/step - loss: 0.1837 - acc: 0.9320 - val_loss: 0.1857 - val_acc: 0.9301\n",
 936 |       "Epoch 10/10\n",
 937 |       "313/313 [==============================] - 3s 9ms/step - loss: 0.1850 - acc: 0.9336\n",
 938 |       "1875/1875 [==============================] - 44s 23ms/step - loss: 0.1757 - acc: 0.9343 - val_loss: 0.1850 - val_acc: 0.9336\n"
 939 |      ]
 940 |     }
 941 |    ],
 942 |    "source": [
 943 |     "# model fitting\n",
 944 |     "history = model.fit_generator(train_generator,\n",
 945 |     "                              epochs=10,\n",
 946 |     "                              verbose=1,\n",
 947 |     "                              validation_data=val_generator, callbacks=callbacks_list)\n",
 948 |     "\n",
 949 |     "# 93.88"
 950 |    ]
 951 |   },
 952 |   {
 953 |    "cell_type": "code",
 954 |    "execution_count": 382,
 955 |    "metadata": {},
 956 |    "outputs": [
 957 |     {
 958 |      "data": {
 959 |       "image/png": "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\n",
 960 |       "text/plain": [
 961 |        "<Figure size 432x288 with 1 Axes>"
 962 |       ]
 963 |      },
 964 |      "metadata": {},
 965 |      "output_type": "display_data"
 966 |     },
 967 |     {
 968 |      "data": {
 969 |       "image/png": "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\n",
 970 |       "text/plain": [
 971 |        "<Figure size 432x288 with 1 Axes>"
 972 |       ]
 973 |      },
 974 |      "metadata": {},
 975 |      "output_type": "display_data"
 976 |     }
 977 |    ],
 978 |    "source": [
 979 |     "# PLOT LOSS AND ACCURACY\n",
 980 |     "import matplotlib.image  as mpimg\n",
 981 |     "import matplotlib.pyplot as plt\n",
 982 |     "\n",
 983 |     "#-----------------------------------------------------------\n",
 984 |     "# Retrieve a list of list results on training and test data\n",
 985 |     "# sets for each training epoch\n",
 986 |     "#-----------------------------------------------------------\n",
 987 |     "acc=history.history['acc']\n",
 988 |     "val_acc=history.history['val_acc']\n",
 989 |     "loss=history.history['loss']\n",
 990 |     "val_loss=history.history['val_loss']\n",
 991 |     "\n",
 992 |     "epochs=range(len(acc)) # Get number of epochs\n",
 993 |     "\n",
 994 |     "#------------------------------------------------\n",
 995 |     "# Plot training and validation accuracy per epoch\n",
 996 |     "#------------------------------------------------\n",
 997 |     "_ = plt.plot(epochs, acc, 'r', \"Training Accuracy\")\n",
 998 |     "_ = plt.plot(epochs, val_acc, 'b', \"Validation Accuracy\")\n",
 999 |     "_ = plt.title('Training and validation accuracy')\n",
1000 |     "_ = plt.figure()\n",
1001 |     "\n",
1002 |     "#------------------------------------------------\n",
1003 |     "# Plot training and validation loss per epoch\n",
1004 |     "#------------------------------------------------\n",
1005 |     "_ = plt.plot(epochs, loss, 'r', \"Training Loss\")\n",
1006 |     "_ = plt.plot(epochs, val_loss, 'b', \"Validation Loss\")\n",
1007 |     "\n",
1008 |     "\n",
1009 |     "_ = plt.title('Training and validation loss')"
1010 |    ]
1011 |   },
1012 |   {
1013 |    "cell_type": "code",
1014 |    "execution_count": 383,
1015 |    "metadata": {},
1016 |    "outputs": [
1017 |     {
1018 |      "name": "stdout",
1019 |      "output_type": "stream",
1020 |      "text": [
1021 |       "predictions shape: (10000, 10)\n"
1022 |      ]
1023 |     }
1024 |    ],
1025 |    "source": [
1026 |     "predictions = model.predict(X_test)\n",
1027 |     "print('predictions shape:', predictions.shape)"
1028 |    ]
1029 |   },
1030 |   {
1031 |    "cell_type": "code",
1032 |    "execution_count": 384,
1033 |    "metadata": {},
1034 |    "outputs": [],
1035 |    "source": [
1036 |     "y_pred = []\n",
1037 |     "for x in predictions:\n",
1038 |     "    y_pred.append(np.argmax(x))"
1039 |    ]
1040 |   },
1041 |   {
1042 |    "cell_type": "code",
1043 |    "execution_count": 385,
1044 |    "metadata": {},
1045 |    "outputs": [
1046 |     {
1047 |      "data": {
1048 |       "text/plain": [
1049 |        "[0, 1, 2, 2, 3]"
1050 |       ]
1051 |      },
1052 |      "execution_count": 385,
1053 |      "metadata": {},
1054 |      "output_type": "execute_result"
1055 |     },
1056 |     {
1057 |      "data": {
1058 |       "text/plain": [
1059 |        "0    0\n",
1060 |        "1    1\n",
1061 |        "2    2\n",
1062 |        "3    2\n",
1063 |        "4    3\n",
1064 |        "Name: label, dtype: int64"
1065 |       ]
1066 |      },
1067 |      "execution_count": 385,
1068 |      "metadata": {},
1069 |      "output_type": "execute_result"
1070 |     }
1071 |    ],
1072 |    "source": [
1073 |     "y_pred[:5]\n",
1074 |     "y_test[:5]"
1075 |    ]
1076 |   },
1077 |   {
1078 |    "cell_type": "code",
1079 |    "execution_count": 386,
1080 |    "metadata": {},
1081 |    "outputs": [
1082 |     {
1083 |      "data": {
1084 |       "text/plain": [
1085 |        "0.9336"
1086 |       ]
1087 |      },
1088 |      "execution_count": 386,
1089 |      "metadata": {},
1090 |      "output_type": "execute_result"
1091 |     }
1092 |    ],
1093 |    "source": [
1094 |     "accuracy_score(y_test, y_pred)"
1095 |    ]
1096 |   },
1097 |   {
1098 |    "cell_type": "code",
1099 |    "execution_count": 387,
1100 |    "metadata": {},
1101 |    "outputs": [],
1102 |    "source": [
1103 |     "from sklearn.metrics import confusion_matrix"
1104 |    ]
1105 |   },
1106 |   {
1107 |    "cell_type": "code",
1108 |    "execution_count": 388,
1109 |    "metadata": {},
1110 |    "outputs": [
1111 |     {
1112 |      "data": {
1113 |       "text/plain": [
1114 |        "array([[864,   1,  27,   8,   0,   0,  98,   0,   2,   0],\n",
1115 |        "       [  0, 994,   1,   2,   0,   0,   3,   0,   0,   0],\n",
1116 |        "       [  9,   0, 894,  10,  45,   0,  42,   0,   0,   0],\n",
1117 |        "       [  7,   7,   6, 936,  22,   0,  21,   0,   1,   0],\n",
1118 |        "       [  0,   0,  26,  10, 923,   0,  40,   0,   1,   0],\n",
1119 |        "       [  0,   0,   0,   0,   0, 966,   0,  25,   3,   6],\n",
1120 |        "       [ 68,   0,  45,  19,  53,   0, 815,   0,   0,   0],\n",
1121 |        "       [  0,   0,   0,   0,   0,   0,   0, 986,   0,  14],\n",
1122 |        "       [  1,   1,   1,   0,   1,   0,   6,   0, 989,   1],\n",
1123 |        "       [  0,   0,   0,   0,   0,   0,   0,  30,   1, 969]])"
1124 |       ]
1125 |      },
1126 |      "execution_count": 388,
1127 |      "metadata": {},
1128 |      "output_type": "execute_result"
1129 |     }
1130 |    ],
1131 |    "source": [
1132 |     "conf_mx = confusion_matrix(y_test, y_pred)\n",
1133 |     "conf_mx"
1134 |    ]
1135 |   },
1136 |   {
1137 |    "cell_type": "code",
1138 |    "execution_count": 389,
1139 |    "metadata": {},
1140 |    "outputs": [
1141 |     {
1142 |      "data": {
1143 |       "text/plain": [
1144 |        "<matplotlib.image.AxesImage at 0x7f6d80917e48>"
1145 |       ]
1146 |      },
1147 |      "execution_count": 389,
1148 |      "metadata": {},
1149 |      "output_type": "execute_result"
1150 |     },
1151 |     {
1152 |      "data": {
1153 |       "image/png": "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\n",
1154 |       "text/plain": [
1155 |        "<Figure size 288x288 with 1 Axes>"
1156 |       ]
1157 |      },
1158 |      "metadata": {},
1159 |      "output_type": "display_data"
1160 |     }
1161 |    ],
1162 |    "source": [
1163 |     "plt.matshow(conf_mx, cmap=plt.cm.gray)\n",
1164 |     "plt.show()"
1165 |    ]
1166 |   },
1167 |   {
1168 |    "cell_type": "code",
1169 |    "execution_count": 390,
1170 |    "metadata": {},
1171 |    "outputs": [],
1172 |    "source": [
1173 |     "row_sums = conf_mx.sum(axis=1, keepdims=True)\n",
1174 |     "norm_conf_mx = conf_mx / row_sums"
1175 |    ]
1176 |   },
1177 |   {
1178 |    "cell_type": "code",
1179 |    "execution_count": 391,
1180 |    "metadata": {},
1181 |    "outputs": [
1182 |     {
1183 |      "data": {
1184 |       "text/plain": [
1185 |        "<matplotlib.image.AxesImage at 0x7f6d808e8f60>"
1186 |       ]
1187 |      },
1188 |      "execution_count": 391,
1189 |      "metadata": {},
1190 |      "output_type": "execute_result"
1191 |     },
1192 |     {
1193 |      "data": {
1194 |       "image/png": "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\n",
1195 |       "text/plain": [
1196 |        "<Figure size 288x288 with 1 Axes>"
1197 |       ]
1198 |      },
1199 |      "metadata": {},
1200 |      "output_type": "display_data"
1201 |     }
1202 |    ],
1203 |    "source": [
1204 |     "np.fill_diagonal(norm_conf_mx, 0)\n",
1205 |     "plt.matshow(norm_conf_mx, cmap=plt.cm.gray)\n",
1206 |     "plt.show()"
1207 |    ]
1208 |   },
1209 |   {
1210 |    "cell_type": "code",
1211 |    "execution_count": 392,
1212 |    "metadata": {},
1213 |    "outputs": [
1214 |     {
1215 |      "data": {
1216 |       "text/plain": [
1217 |        "'\\n0 T-shirt/top\\n1 Trouser\\n2 Pullover\\n3 Dress\\n4 Coat\\n5 Sandal\\n6 Shirt\\n7 Sneaker\\n8 Bag\\n9 Ankle boot\\n'"
1218 |       ]
1219 |      },
1220 |      "execution_count": 392,
1221 |      "metadata": {},
1222 |      "output_type": "execute_result"
1223 |     }
1224 |    ],
1225 |    "source": [
1226 |     "'''\n",
1227 |     "0 T-shirt/top\n",
1228 |     "1 Trouser\n",
1229 |     "2 Pullover\n",
1230 |     "3 Dress\n",
1231 |     "4 Coat\n",
1232 |     "5 Sandal\n",
1233 |     "6 Shirt\n",
1234 |     "7 Sneaker\n",
1235 |     "8 Bag\n",
1236 |     "9 Ankle boot\n",
1237 |     "'''"
1238 |    ]
1239 |   },
1240 |   {
1241 |    "cell_type": "code",
1242 |    "execution_count": null,
1243 |    "metadata": {},
1244 |    "outputs": [],
1245 |    "source": []
1246 |   }
1247 |  ],
1248 |  "metadata": {
1249 |   "kernelspec": {
1250 |    "display_name": "Environment (conda_tensorflow_p36)",
1251 |    "language": "python",
1252 |    "name": "conda_tensorflow_p36"
1253 |   },
1254 |   "language_info": {
1255 |    "codemirror_mode": {
1256 |     "name": "ipython",
1257 |     "version": 3
1258 |    },
1259 |    "file_extension": ".py",
1260 |    "mimetype": "text/x-python",
1261 |    "name": "python",
1262 |    "nbconvert_exporter": "python",
1263 |    "pygments_lexer": "ipython3",
1264 |    "version": "3.6.10"
1265 |   }
1266 |  },
1267 |  "nbformat": 4,
1268 |  "nbformat_minor": 2
1269 | }
1270 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
1 | # Kaggle-notebooks


--------------------------------------------------------------------------------