Ce serveur Gitlab sera éteint le 30 juin 2020, pensez à migrer vos projets vers les serveurs gitlab-research.centralesupelec.fr et gitlab-student.centralesupelec.fr !

Spectrum_Keras-Multi-Cat.ipynb 108 KB
Newer Older
SoleneDc's avatar
SoleneDc committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. CNN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Récupération des genres"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Permet de récupérer les labels qui seront mis dans une array"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
26
   "execution_count": 27,
SoleneDc's avatar
SoleneDc committed
27 28 29 30 31 32 33 34 35
   "metadata": {},
   "outputs": [],
   "source": [
    "import ast\n",
    "import pandas as pd\n",
    "from __future__ import absolute_import\n",
    "from __future__ import division\n",
    "from __future__ import print_function\n",
    "from keras.preprocessing.sequence import pad_sequences\n",
SoleneDc's avatar
SoleneDc committed
36
    "pd.options.mode.chained_assignment = None\n",
SoleneDc's avatar
SoleneDc committed
37 38 39 40 41 42 43 44 45 46 47 48
    "import argparse\n",
    "import sys\n",
    "import numpy as np\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.image as mpimg\n",
    "import tensorflow as tf\n",
    "import os\n",
    "import cv2\n",
    "from math import floor"
   ]
  },
SoleneDc's avatar
SoleneDc committed
49 50 51 52 53 54 55
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Choisir l'année !"
   ]
  },
SoleneDc's avatar
SoleneDc committed
56 57
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
58 59 60 61 62 63 64 65 66 67 68 69
   "execution_count": 86,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "year = 2007"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
SoleneDc's avatar
SoleneDc committed
70 71 72 73 74 75
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "list_of_eligible_spectrums = []\n",
SoleneDc's avatar
SoleneDc committed
76
    "for file in os.listdir(\"spectrumImages/SpectrumImages\" + str(year)):\n",
SoleneDc's avatar
SoleneDc committed
77 78 79 80
    "    if str(file)[-4:] == '.jpg':\n",
    "        list_of_eligible_spectrums += [file]"
   ]
  },
SoleneDc's avatar
SoleneDc committed
81 82 83 84 85 86 87
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
88 89 90 91 92 93 94
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "{\"genres\":[{\"id\":28,\"name\":\"Action\"},{\"id\":12,\"name\":\"Adventure\"},{\"id\":16,\"name\":\"Animation\"},{\"id\":35,\"name\":\"Comedy\"},{\"id\":80,\"name\":\"Crime\"},{\"id\":99,\"name\":\"Documentary\"},{\"id\":18,\"name\":\"Drama\"},{\"id\":10751,\"name\":\"Family\"},{\"id\":14,\"name\":\"Fantasy\"},{\"id\":36,\"name\":\"History\"},{\"id\":27,\"name\":\"Horror\"},{\"id\":10402,\"name\":\"Music\"},{\"id\":9648,\"name\":\"Mystery\"},{\"id\":10749,\"name\":\"Romance\"},{\"id\":878,\"name\":\"Science Fiction\"},{\"id\":10770,\"name\":\"TV Movie\"},{\"id\":53,\"name\":\"Thriller\"},{\"id\":10752,\"name\":\"War\"},{\"id\":37,\"name\":\"Western\"}]}"
   ]
  },
SoleneDc's avatar
SoleneDc committed
95 96
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
97 98
   "execution_count": 89,
   "metadata": {},
SoleneDc's avatar
SoleneDc committed
99 100 101 102 103
   "outputs": [],
   "source": [
    "genres = [28, 35, 18, 99, 10749, 10752, 10402, 53, 878, 27, 9648, 80, 14, 12, 36, 10769, 16, 10751, 37, 10770]\n",
    "\n",
    "def get_genre_from_link():\n",
SoleneDc's avatar
SoleneDc committed
104
    "    path = \"./Link-dictionaries/Link-dictionary\" + str(year)+ \".txt\"\n",
SoleneDc's avatar
SoleneDc committed
105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140
    "    file = open(path, \"r\").read()\n",
    "    dictyear = ast.literal_eval(file)\n",
    "    dict_inverse = {}\n",
    "    links_to_be_removed = []\n",
    "    for movie_id in dictyear.keys():\n",
    "        if dictyear[movie_id][1] != []:\n",
    "            dict_inverse[str(dictyear[movie_id][2])] = {}\n",
    "            for genre in genres:\n",
    "                if genre in dictyear[movie_id][1]:\n",
    "                    dict_inverse[str(dictyear[movie_id][2])][genre] = 1\n",
    "                else:\n",
    "                    dict_inverse[str(dictyear[movie_id][2])][genre] = 0\n",
    "        else:\n",
    "            #print(f'careful, link {dictyear[movie_id][2]} needs to be removed from the list')\n",
    "            links_to_be_removed += [dictyear[movie_id][2]]\n",
    "    return dict_inverse, links_to_be_removed\n",
    "\n",
    "def get_output_list(L):\n",
    "    dict_inverse, links_to_be_removed = get_genre_from_link()\n",
    "    eligible_links = []\n",
    "    output = []\n",
    "    for link in L:\n",
    "        link = str(link)\n",
    "        #print(dict_inverse[str(link)])\n",
    "        if link[-1] == \".\":\n",
    "            print(\"do something! Too many points.......\")\n",
    "        if link[:-4] not in links_to_be_removed:\n",
    "            eligible_links += [link[:-4]]\n",
    "    return dict_inverse, eligible_links\n",
    "\n",
    "\n",
    "dict_inverse, eligible_links = get_output_list(list_of_eligible_spectrums)"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
141 142 143 144
   "execution_count": 90,
   "metadata": {
    "collapsed": true
   },
SoleneDc's avatar
SoleneDc committed
145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182
   "outputs": [],
   "source": [
    "#for element in labels:\n",
    "#   for genre in element:\n",
    "#       if genre not in genres:\n",
    "#           genres += [genre]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#Modif pour ajouter des catégories\n",
    "trY[2][3]=1\n",
    "trY.shape\n",
    "from random import randint\n",
    "for i in range(1225):\n",
    "    rand = randint(0,19)\n",
    "    trY[i][rand] = 1\n",
    "trY[3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Bien vérifier la taille des données !"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Extraction des images"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
183
   "execution_count": 92,
SoleneDc's avatar
SoleneDc committed
184 185 186 187
   "metadata": {},
   "outputs": [],
   "source": [
    "for file in eligible_links:\n",
SoleneDc's avatar
SoleneDc committed
188
    "    img = cv2.imread('SpectrumImages/SpectrumImages' + str(year) + '/' + file + '.jpg', 1)\n",
SoleneDc's avatar
SoleneDc committed
189 190 191 192 193 194 195
    "    img = img[0:1]\n",
    "    img = img.reshape((img.shape[1], img.shape[2]))\n",
    "    dict_inverse[file]['image'] = img"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
196 197 198 199
   "execution_count": 93,
   "metadata": {
    "collapsed": true
   },
SoleneDc's avatar
SoleneDc committed
200 201 202 203 204 205 206 207 208 209 210
   "outputs": [],
   "source": [
    "df = pd.DataFrame.from_dict(dict_inverse)\n",
    "df = df.transpose()\n",
    "df = df.reset_index(drop=True)\n",
    "#shuffling\n",
    "df = df.sample(frac=1)"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
211 212 213 214
   "execution_count": 94,
   "metadata": {
    "collapsed": true
   },
SoleneDc's avatar
SoleneDc committed
215 216 217 218 219 220 221
   "outputs": [],
   "source": [
    "df2 = df.dropna(axis=0)"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
222
   "execution_count": 95,
SoleneDc's avatar
SoleneDc committed
223 224 225 226 227
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
228
       "(1910, 21)"
SoleneDc's avatar
SoleneDc committed
229 230
      ]
     },
SoleneDc's avatar
SoleneDc committed
231
     "execution_count": 95,
SoleneDc's avatar
SoleneDc committed
232 233 234 235 236 237 238 239 240 241
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
242
   "execution_count": 96,
SoleneDc's avatar
SoleneDc committed
243 244 245 246 247
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
248
       "(1612, 21)"
SoleneDc's avatar
SoleneDc committed
249 250
      ]
     },
SoleneDc's avatar
SoleneDc committed
251
     "execution_count": 96,
SoleneDc's avatar
SoleneDc committed
252 253 254 255 256 257 258 259 260 261
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.shape"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
262
   "execution_count": 97,
263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>12</th>\n",
       "      <th>14</th>\n",
       "      <th>16</th>\n",
       "      <th>18</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>53</th>\n",
       "      <th>...</th>\n",
       "      <th>99</th>\n",
       "      <th>878</th>\n",
       "      <th>9648</th>\n",
       "      <th>10402</th>\n",
       "      <th>10749</th>\n",
       "      <th>10751</th>\n",
       "      <th>10752</th>\n",
       "      <th>10769</th>\n",
       "      <th>10770</th>\n",
       "      <th>image</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
SoleneDc's avatar
SoleneDc committed
312 313 314 315 316 317 318 319 320 321
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
322
       "      <td>...</td>\n",
SoleneDc's avatar
SoleneDc committed
323 324 325 326 327 328 329 330 331 332
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
       "      <td>1612</td>\n",
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
SoleneDc's avatar
SoleneDc committed
356
       "      <td>1612</td>\n",
357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
380
       "      <td>[[62, 146, 44], [62, 146, 44], [62, 146, 44], ...</td>\n",
381 382 383
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
SoleneDc's avatar
SoleneDc committed
384 385 386 387 388 389 390 391 392 393
       "      <td>1465</td>\n",
       "      <td>1499</td>\n",
       "      <td>1496</td>\n",
       "      <td>870</td>\n",
       "      <td>1373</td>\n",
       "      <td>1340</td>\n",
       "      <td>1177</td>\n",
       "      <td>1559</td>\n",
       "      <td>1589</td>\n",
       "      <td>1263</td>\n",
394
       "      <td>...</td>\n",
SoleneDc's avatar
SoleneDc committed
395 396 397 398 399 400 401 402 403
       "      <td>1413</td>\n",
       "      <td>1507</td>\n",
       "      <td>1522</td>\n",
       "      <td>1538</td>\n",
       "      <td>1364</td>\n",
       "      <td>1486</td>\n",
       "      <td>1573</td>\n",
       "      <td>1560</td>\n",
       "      <td>1588</td>\n",
404 405 406 407 408 409 410 411 412
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          12    14    16    18    27    28    35    36    37    53  \\\n",
SoleneDc's avatar
SoleneDc committed
413
       "count   1612  1612  1612  1612  1612  1612  1612  1612  1612  1612   \n",
414 415
       "unique     2     2     2     2     2     2     2     2     2     2   \n",
       "top        0     0     0     0     0     0     0     0     0     0   \n",
SoleneDc's avatar
SoleneDc committed
416
       "freq    1465  1499  1496   870  1373  1340  1177  1559  1589  1263   \n",
417 418
       "\n",
       "                              ...                            99   878  9648  \\\n",
SoleneDc's avatar
SoleneDc committed
419
       "count                         ...                          1612  1612  1612   \n",
420 421
       "unique                        ...                             2     2     2   \n",
       "top                           ...                             0     0     0   \n",
SoleneDc's avatar
SoleneDc committed
422
       "freq                          ...                          1413  1507  1522   \n",
423 424
       "\n",
       "        10402  10749  10751  10752  10769  10770  \\\n",
SoleneDc's avatar
SoleneDc committed
425
       "count    1612   1612   1612   1612   1612   1612   \n",
426 427
       "unique      2      2      2      2      2      2   \n",
       "top         0      0      0      0      0      0   \n",
SoleneDc's avatar
SoleneDc committed
428
       "freq     1538   1364   1486   1573   1560   1588   \n",
429 430
       "\n",
       "                                                    image  \n",
SoleneDc's avatar
SoleneDc committed
431 432 433
       "count                                                1612  \n",
       "unique                                               1612  \n",
       "top     [[62, 146, 44], [62, 146, 44], [62, 146, 44], ...  \n",
434 435 436 437 438
       "freq                                                    1  \n",
       "\n",
       "[4 rows x 21 columns]"
      ]
     },
SoleneDc's avatar
SoleneDc committed
439
     "execution_count": 97,
440 441 442 443 444 445 446 447 448 449
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.describe()"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>12</th>\n",
       "      <th>14</th>\n",
       "      <th>16</th>\n",
       "      <th>18</th>\n",
       "      <th>27</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>53</th>\n",
       "      <th>80</th>\n",
       "      <th>99</th>\n",
       "      <th>878</th>\n",
       "      <th>9648</th>\n",
       "      <th>10402</th>\n",
       "      <th>10749</th>\n",
       "      <th>10751</th>\n",
       "      <th>10752</th>\n",
       "      <th>10769</th>\n",
       "      <th>10770</th>\n",
       "      <th>image</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "      <td>1340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "      <td>272</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      12    14    16    18    27    35    36    37    53    80    99   878  \\\n",
       "28                                                                           \n",
       "0   1340  1340  1340  1340  1340  1340  1340  1340  1340  1340  1340  1340   \n",
       "1    272   272   272   272   272   272   272   272   272   272   272   272   \n",
       "\n",
       "    9648  10402  10749  10751  10752  10769  10770  image  \n",
       "28                                                         \n",
       "0   1340   1340   1340   1340   1340   1340   1340   1340  \n",
       "1    272    272    272    272    272    272    272    272  "
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.groupby(28).count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 61,
SoleneDc's avatar
SoleneDc committed
601 602 603 604 605
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
606
       "1289"
SoleneDc's avatar
SoleneDc committed
607 608
      ]
     },
SoleneDc's avatar
SoleneDc committed
609
     "execution_count": 61,
SoleneDc's avatar
SoleneDc committed
610 611 612 613 614
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
615
    "train_len = int(df2.shape[0]*0.8)\n",
SoleneDc's avatar
SoleneDc committed
616 617 618 619 620
    "train_len"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
621 622 623 624
   "execution_count": 62,
   "metadata": {
    "collapsed": true
   },
SoleneDc's avatar
SoleneDc committed
625 626 627 628 629 630 631 632
   "outputs": [],
   "source": [
    "train = df2.iloc[:train_len, :]\n",
    "test = df2.iloc[train_len:, :]"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
633
   "execution_count": 63,
SoleneDc's avatar
SoleneDc committed
634 635 636 637 638
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
639 640 641 642 643
       "1152    [[14, 6, 0], [16, 6, 0], [16, 5, 1], [16, 6, 0...\n",
       "1019    [[13, 0, 9], [0, 13, 0], [49, 114, 39], [38, 1...\n",
       "992     [[5, 5, 0], [30, 25, 16], [24, 17, 2], [20, 15...\n",
       "983     [[85, 85, 49], [105, 106, 67], [88, 91, 52], [...\n",
       "603     [[19, 16, 1], [37, 33, 22], [42, 37, 28], [37,...\n",
SoleneDc's avatar
SoleneDc committed
644 645 646
       "Name: image, dtype: object"
      ]
     },
SoleneDc's avatar
SoleneDc committed
647
     "execution_count": 63,
SoleneDc's avatar
SoleneDc committed
648 649 650 651 652 653 654 655 656 657
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['image'].head()"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
658
   "execution_count": 64,
SoleneDc's avatar
SoleneDc committed
659 660 661 662 663
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
664
       "((1289, 21), (323, 21))"
SoleneDc's avatar
SoleneDc committed
665 666
      ]
     },
SoleneDc's avatar
SoleneDc committed
667
     "execution_count": 64,
SoleneDc's avatar
SoleneDc committed
668 669 670 671 672 673 674 675 676 677
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape, test.shape"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
678
   "execution_count": 65,
SoleneDc's avatar
SoleneDc committed
679
   "metadata": {},
SoleneDc's avatar
SoleneDc committed
680
   "outputs": [],
SoleneDc's avatar
SoleneDc committed
681
   "source": [
682 683 684 685
    "selected_genre = 28\n",
    "name = 'not' + str(selected_genre)\n",
    "train[name] = np.where(train[selected_genre] == 1, 0, 1)\n",
    "test[name] = np.where(test[selected_genre] == 1, 0, 1)\n",
SoleneDc's avatar
SoleneDc committed
686
    "X_train = train['image']\n",
687 688
    "Y_train = train[[selected_genre, name]]\n",
    "#Y_train = train.drop('image', 1)\n",
SoleneDc's avatar
SoleneDc committed
689
    "X_test = test['image']\n",
690 691
    "Y_test = test[[selected_genre, name]]\n",
    "#Y_test = test.drop('image', 1)"
SoleneDc's avatar
SoleneDc committed
692 693 694 695 696
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
SoleneDc's avatar
SoleneDc committed
697 698 699
   "metadata": {
    "collapsed": true
   },
SoleneDc's avatar
SoleneDc committed
700
   "outputs": [],
701 702 703
   "source": [
    "\n"
   ]
SoleneDc's avatar
SoleneDc committed
704 705 706
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
707
   "execution_count": 66,
SoleneDc's avatar
SoleneDc committed
708 709 710 711
   "metadata": {},
   "outputs": [
    {
     "data": {
712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>12</th>\n",
       "      <th>14</th>\n",
       "      <th>16</th>\n",
       "      <th>18</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>53</th>\n",
       "      <th>...</th>\n",
       "      <th>878</th>\n",
       "      <th>9648</th>\n",
       "      <th>10402</th>\n",
       "      <th>10749</th>\n",
       "      <th>10751</th>\n",
       "      <th>10752</th>\n",
       "      <th>10769</th>\n",
       "      <th>10770</th>\n",
       "      <th>image</th>\n",
SoleneDc's avatar
SoleneDc committed
751
       "      <th>not28</th>\n",
752 753 754 755
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
SoleneDc's avatar
SoleneDc committed
756
       "      <th>1152</th>\n",
757 758 759 760 761 762 763 764 765 766
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
767
       "      <td>...</td>\n",
768 769 770 771 772 773 774 775
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
776 777
       "      <td>[[14, 6, 0], [16, 6, 0], [16, 5, 1], [16, 6, 0...</td>\n",
       "      <td>1</td>\n",
778 779
       "    </tr>\n",
       "    <tr>\n",
SoleneDc's avatar
SoleneDc committed
780
       "      <th>1019</th>\n",
781 782 783
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
784
       "      <td>1</td>\n",
785 786 787 788 789
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
790
       "      <td>1</td>\n",
791 792 793 794 795 796 797 798 799
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
800 801
       "      <td>[[13, 0, 9], [0, 13, 0], [49, 114, 39], [38, 1...</td>\n",
       "      <td>0</td>\n",
802 803
       "    </tr>\n",
       "    <tr>\n",
SoleneDc's avatar
SoleneDc committed
804
       "      <th>992</th>\n",
805 806 807
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
808
       "      <td>1</td>\n",
809
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
810
       "      <td>1</td>\n",
811 812 813 814 815 816 817 818 819 820 821 822 823
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
824
       "      <td>[[5, 5, 0], [30, 25, 16], [24, 17, 2], [20, 15...</td>\n",
825 826 827
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
SoleneDc's avatar
SoleneDc committed
828 829 830
       "      <th>983</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
831 832 833 834 835 836 837 838 839 840 841 842 843 844 845
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
846 847 848
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>[[85, 85, 49], [105, 106, 67], [88, 91, 52], [...</td>\n",
849 850 851
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
SoleneDc's avatar
SoleneDc committed
852
       "      <th>603</th>\n",
853 854 855 856 857 858 859 860 861 862
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
863 864
       "      <td>...</td>\n",
       "      <td>0</td>\n",
865 866 867 868 869 870 871
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
SoleneDc's avatar
SoleneDc committed
872 873
       "      <td>[[19, 16, 1], [37, 33, 22], [42, 37, 28], [37,...</td>\n",
       "      <td>1</td>\n",
874 875 876 877 878 879
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
SoleneDc's avatar
SoleneDc committed
880
      "text/plain": [
881
       "     12 14 16 18 27 28 35 36 37 53  ...  878 9648 10402 10749 10751 10752  \\\n",
SoleneDc's avatar
SoleneDc committed
882 883 884 885 886
       "1152  0  0  0  1  0  0  0  0  0  0  ...    0    0     0     0     0     0   \n",
       "1019  0  0  0  1  0  1  0  0  0  1  ...    0    0     0     0     0     0   \n",
       "992   0  0  0  1  0  1  0  0  0  0  ...    0    0     0     0     0     0   \n",
       "983   0  0  0  0  0  0  1  0  0  0  ...    0    0     0     0     0     0   \n",
       "603   0  0  0  1  0  0  0  0  0  0  ...    0    0     0     0     0     1   \n",
887
       "\n",
SoleneDc's avatar
SoleneDc committed
888 889 890 891 892 893
       "     10769 10770                                              image not28  \n",
       "1152     0     0  [[14, 6, 0], [16, 6, 0], [16, 5, 1], [16, 6, 0...     1  \n",
       "1019     0     0  [[13, 0, 9], [0, 13, 0], [49, 114, 39], [38, 1...     0  \n",
       "992      0     0  [[5, 5, 0], [30, 25, 16], [24, 17, 2], [20, 15...     0  \n",
       "983      0     0  [[85, 85, 49], [105, 106, 67], [88, 91, 52], [...     1  \n",
       "603      0     0  [[19, 16, 1], [37, 33, 22], [42, 37, 28], [37,...     1  \n",
894 895
       "\n",
       "[5 rows x 22 columns]"
SoleneDc's avatar
SoleneDc committed
896 897
      ]
     },
SoleneDc's avatar
SoleneDc committed
898
     "execution_count": 66,
SoleneDc's avatar
SoleneDc committed
899 900 901 902 903
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
904 905 906 907 908
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
909
   "execution_count": 67,
910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
SoleneDc's avatar
SoleneDc committed
933
       "      <th>not28</th>\n",
934 935
       "    </tr>\n",
       "    <tr>\n",
SoleneDc's avatar
SoleneDc committed
936
       "      <th>28</th>\n",
937 938 939 940 941 942
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
SoleneDc's avatar
SoleneDc committed
943
       "      <td>1080</td>\n",
944 945 946
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
SoleneDc's avatar
SoleneDc committed
947
       "      <td>209</td>\n",
948 949 950 951 952 953
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
954 955 956 957
       "    not28\n",
       "28       \n",
       "0    1080\n",
       "1     209"
958 959
      ]
     },
SoleneDc's avatar
SoleneDc committed
960
     "execution_count": 67,
961 962 963 964 965 966 967 968 969 970
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y_train.groupby(selected_genre).count()"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
971
   "execution_count": 81,
972 973 974 975 976
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
977
       "0.8378588052754073"
978 979
      ]
     },
SoleneDc's avatar
SoleneDc committed
980
     "execution_count": 81,
981 982 983 984 985
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
SoleneDc's avatar
SoleneDc committed
986
    "1080/(1080+209)"
SoleneDc's avatar
SoleneDc committed
987 988 989 990 991 992 993 994 995 996 997
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Vérifications des données"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
998 999 1000 1001
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
SoleneDc's avatar
SoleneDc committed
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
1014 1015 1016 1017
   "execution_count": 69,
   "metadata": {
    "collapsed": true
   },
SoleneDc's avatar
SoleneDc committed
1018 1019 1020 1021 1022 1023 1024 1025
   "outputs": [],
   "source": [
    "X_train = pad_sequences(X_train)\n",
    "X_test = pad_sequences(X_test)"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
1026 1027
   "execution_count": 105,
   "metadata": {},
SoleneDc's avatar
SoleneDc committed
1028
   "outputs": [],
SoleneDc's avatar
SoleneDc committed
1029 1030 1031
   "source": [
    "input_shapeA = (X_train.shape[1], X_train.shape[2])"
   ]
SoleneDc's avatar
SoleneDc committed
1032 1033 1034 1035 1036 1037 1038 1039 1040 1041
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modèle"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
1042
   "execution_count": 106,
SoleneDc's avatar
SoleneDc committed
1043 1044 1045 1046 1047 1048
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
SoleneDc's avatar
SoleneDc committed
1049 1050 1051 1052 1053
      "Conv1D 1 : (None, 1198, 2)\n",
      "MaxP1D 1 : (None, 599, 2)\n",
      "Conv1D 2 : (None, 597, 4)\n",
      "MaxP1D 2 : (None, 298, 4)\n",
      "Flatten : (None, 1192)\n",
1054
      "Dense  2 : (None, 2)\n"
SoleneDc's avatar
SoleneDc committed
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
     ]
    }
   ],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation\n",
    "from keras.layers import Conv1D, MaxPooling1D\n",
    "from keras.layers import Dropout, Average, BatchNormalization\n",
    "from keras.layers import Flatten\n",
    "\n",
1065
    "num_classes=2\n",
SoleneDc's avatar
SoleneDc committed
1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079
    "\n",
    "#Hyperparameters\n",
    "filtersCNN1=2 #pourquoi ??\n",
    "kernelSize1=3\n",
    "\n",
    "filtersCNN2=4\n",
    "kernelSize2=3\n",
    "\n",
    "unitsFC1=1000\n",
    "unitsFC2=num_classes\n",
    "\n",
    "#defining the layers architecture\n",
    "\n",
    "model = Sequential()\n",
SoleneDc's avatar
SoleneDc committed
1080
    "model.add(Conv1D(filtersCNN1,kernelSize1,strides=1, padding=\"valid\", activation='relu',input_shape=input_shapeA))\n",
SoleneDc's avatar
SoleneDc committed
1081 1082 1083 1084
    "print(\"Conv1D 1 : {}\".format(model.output_shape))\n",
    "model.add(MaxPooling1D(pool_size=2,padding=\"valid\"))\n",
    "print(\"MaxP1D 1 : {}\".format(model.output_shape))\n",
    "#BatchNormalization(axis=3)\n",
SoleneDc's avatar
SoleneDc committed
1085
    "model.add(Dropout(0.25))\n",
SoleneDc's avatar
SoleneDc committed
1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102
    "\n",
    "model.add(Conv1D(filtersCNN2,kernelSize2,strides=1, padding=\"valid\", activation='relu'))\n",
    "print(\"Conv1D 2 : {}\".format(model.output_shape))\n",
    "model.add(MaxPooling1D(pool_size=2,strides=None,padding=\"valid\"))\n",
    "print(\"MaxP1D 2 : {}\".format(model.output_shape))\n",
    "#BatchNormalization(axis=3)\n",
    "\n",
    "model.add(Flatten())\n",
    "print(\"Flatten : {}\".format(model.output_shape))\n",
    "#model.add(Dense(1000, activation='relu'))\n",
    "#print(\"Dense  1 : {}\".format(model.output_shape))\n",
    "model.add(Dense(num_classes, activation='softmax'))\n",
    "print(\"Dense  2 : {}\".format(model.output_shape))"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
1103
   "execution_count": 107,
SoleneDc's avatar
SoleneDc committed
1104 1105 1106 1107 1108 1109 1110 1111 1112
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
SoleneDc's avatar
SoleneDc committed
1113 1114 1115
      "conv1d_13 (Conv1D)           (None, 1198, 2)           20        \n",
      "_________________________________________________________________\n",
      "max_pooling1d_13 (MaxPooling (None, 599, 2)            0         \n",
SoleneDc's avatar
SoleneDc committed
1116
      "_________________________________________________________________\n",
SoleneDc's avatar
SoleneDc committed
1117
      "dropout_2 (Dropout)          (None, 599, 2)            0         \n",
SoleneDc's avatar
SoleneDc committed
1118
      "_________________________________________________________________\n",
SoleneDc's avatar
SoleneDc committed
1119
      "conv1d_14 (Conv1D)           (None, 597, 4)            28        \n",
SoleneDc's avatar
SoleneDc committed
1120
      "_________________________________________________________________\n",
SoleneDc's avatar
SoleneDc committed
1121
      "max_pooling1d_14 (MaxPooling (None, 298, 4)            0         \n",
SoleneDc's avatar
SoleneDc committed
1122
      "_________________________________________________________________\n",
SoleneDc's avatar
SoleneDc committed
1123
      "flatten_7 (Flatten)          (None, 1192)              0         \n",
SoleneDc's avatar
SoleneDc committed
1124
      "_________________________________________________________________\n",
SoleneDc's avatar
SoleneDc committed
1125
      "dense_7 (Dense)              (None, 2)                 2386      \n",
SoleneDc's avatar
SoleneDc committed
1126
      "=================================================================\n",
SoleneDc's avatar
SoleneDc committed
1127 1128
      "Total params: 2,434\n",
      "Trainable params: 2,434\n",
SoleneDc's avatar
SoleneDc committed
1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Entrainement du modèle"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
1147
   "execution_count": 108,
SoleneDc's avatar
SoleneDc committed
1148 1149 1150 1151 1152 1153
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
SoleneDc's avatar
SoleneDc committed
1154
      "Train on 902 samples, validate on 387 samples\n",
SoleneDc's avatar
SoleneDc committed
1155
      "Epoch 1/100\n",
SoleneDc's avatar
SoleneDc committed
1156
      "902/902 [==============================] - 1s 1ms/step - loss: 3.0958 - acc: 0.7517 - val_loss: 2.3489 - val_acc: 0.8424\n",
SoleneDc's avatar
SoleneDc committed
1157
      "Epoch 2/100\n",
SoleneDc's avatar
SoleneDc committed
1158
      "902/902 [==============================] - 0s 340us/step - loss: 2.5514 - acc: 0.8282 - val_loss: 2.3181 - val_acc: 0.8372\n",
SoleneDc's avatar
SoleneDc committed
1159
      "Epoch 3/100\n",
SoleneDc's avatar
SoleneDc committed
1160
      "902/902 [==============================] - 0s 378us/step - loss: 2.4956 - acc: 0.8182 - val_loss: 2.2628 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1161
      "Epoch 4/100\n",
SoleneDc's avatar
SoleneDc committed
1162
      "902/902 [==============================] - 0s 357us/step - loss: 2.3319 - acc: 0.8071 - val_loss: 2.1581 - val_acc: 0.7959\n",
SoleneDc's avatar
SoleneDc committed
1163
      "Epoch 5/100\n",
SoleneDc's avatar
SoleneDc committed
1164
      "902/902 [==============================] - 0s 332us/step - loss: 1.9852 - acc: 0.7838 - val_loss: 1.8762 - val_acc: 0.7597\n",
SoleneDc's avatar
SoleneDc committed
1165
      "Epoch 6/100\n",
SoleneDc's avatar
SoleneDc committed
1166
      "902/902 [==============================] - 0s 332us/step - loss: 1.5901 - acc: 0.7328 - val_loss: 1.4140 - val_acc: 0.7364\n",
SoleneDc's avatar
SoleneDc committed
1167
      "Epoch 7/100\n",
SoleneDc's avatar
SoleneDc committed
1168
      "902/902 [==============================] - 0s 319us/step - loss: 1.3478 - acc: 0.7373 - val_loss: 1.0776 - val_acc: 0.6925\n",
SoleneDc's avatar
SoleneDc committed
1169
      "Epoch 8/100\n",
SoleneDc's avatar
SoleneDc committed
1170
      "902/902 [==============================] - 0s 379us/step - loss: 0.9999 - acc: 0.7406 - val_loss: 0.9283 - val_acc: 0.7571\n",
SoleneDc's avatar
SoleneDc committed
1171
      "Epoch 9/100\n",
SoleneDc's avatar
SoleneDc committed
1172
      "902/902 [==============================] - 0s 391us/step - loss: 0.8553 - acc: 0.7860 - val_loss: 0.8334 - val_acc: 0.7390\n",
SoleneDc's avatar
SoleneDc committed
1173
      "Epoch 10/100\n",
SoleneDc's avatar
SoleneDc committed
1174
      "902/902 [==============================] - 0s 304us/step - loss: 0.8022 - acc: 0.7816 - val_loss: 0.7862 - val_acc: 0.7700\n",
SoleneDc's avatar
SoleneDc committed
1175
      "Epoch 11/100\n",
SoleneDc's avatar
SoleneDc committed
1176
      "902/902 [==============================] - 0s 459us/step - loss: 0.7167 - acc: 0.7993 - val_loss: 0.7058 - val_acc: 0.8088\n",
SoleneDc's avatar
SoleneDc committed
1177
      "Epoch 12/100\n",
SoleneDc's avatar
SoleneDc committed
1178
      "902/902 [==============================] - 0s 380us/step - loss: 0.6379 - acc: 0.8215 - val_loss: 0.6045 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1179
      "Epoch 13/100\n",
SoleneDc's avatar
SoleneDc committed
1180
      "902/902 [==============================] - 0s 310us/step - loss: 0.4992 - acc: 0.8381 - val_loss: 0.5958 - val_acc: 0.8191\n",
SoleneDc's avatar
SoleneDc committed
1181
      "Epoch 14/100\n",
SoleneDc's avatar
SoleneDc committed
1182
      "902/902 [==============================] - 0s 392us/step - loss: 0.5222 - acc: 0.8248 - val_loss: 0.5783 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1183
      "Epoch 15/100\n",
SoleneDc's avatar
SoleneDc committed
1184
      "902/902 [==============================] - 0s 303us/step - loss: 0.5465 - acc: 0.8226 - val_loss: 0.7402 - val_acc: 0.7623\n",
SoleneDc's avatar
SoleneDc committed
1185
      "Epoch 16/100\n",
SoleneDc's avatar
SoleneDc committed
1186
      "902/902 [==============================] - 0s 344us/step - loss: 0.5189 - acc: 0.8315 - val_loss: 0.5925 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1187
      "Epoch 17/100\n",
SoleneDc's avatar
SoleneDc committed
1188
      "902/902 [==============================] - 0s 292us/step - loss: 0.4669 - acc: 0.8348 - val_loss: 0.5782 - val_acc: 0.8424\n",
SoleneDc's avatar
SoleneDc committed
1189
      "Epoch 18/100\n",
SoleneDc's avatar
SoleneDc committed
1190
      "902/902 [==============================] - 0s 296us/step - loss: 0.4446 - acc: 0.8359 - val_loss: 0.5598 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1191
      "Epoch 19/100\n",
SoleneDc's avatar
SoleneDc committed
1192
      "902/902 [==============================] - 0s 342us/step - loss: 0.4647 - acc: 0.8381 - val_loss: 0.6331 - val_acc: 0.7984\n",
SoleneDc's avatar
SoleneDc committed
1193
      "Epoch 20/100\n",
SoleneDc's avatar
SoleneDc committed
1194
      "902/902 [==============================] - 0s 289us/step - loss: 0.4732 - acc: 0.8215 - val_loss: 0.6496 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1195
      "Epoch 21/100\n",
SoleneDc's avatar
SoleneDc committed
1196
      "902/902 [==============================] - 0s 320us/step - loss: 0.4640 - acc: 0.8459 - val_loss: 0.6247 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1197
      "Epoch 22/100\n",
SoleneDc's avatar
SoleneDc committed
1198
      "902/902 [==============================] - 0s 302us/step - loss: 0.4167 - acc: 0.8404 - val_loss: 0.6707 - val_acc: 0.8010\n",
SoleneDc's avatar
SoleneDc committed
1199
      "Epoch 23/100\n",
SoleneDc's avatar
SoleneDc committed
1200
      "902/902 [==============================] - 0s 467us/step - loss: 0.4106 - acc: 0.8503 - val_loss: 0.6108 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1201
      "Epoch 24/100\n",
SoleneDc's avatar
SoleneDc committed
1202
      "902/902 [==============================] - 0s 411us/step - loss: 0.4243 - acc: 0.8459 - val_loss: 0.6171 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1203
      "Epoch 25/100\n",
SoleneDc's avatar
SoleneDc committed
1204
      "902/902 [==============================] - 0s 440us/step - loss: 0.4298 - acc: 0.8437 - val_loss: 0.6059 - val_acc: 0.8372\n",
SoleneDc's avatar
SoleneDc committed
1205
      "Epoch 26/100\n",
SoleneDc's avatar
SoleneDc committed
1206
      "902/902 [==============================] - 0s 449us/step - loss: 0.4276 - acc: 0.8525 - val_loss: 0.5988 - val_acc: 0.8398\n",
SoleneDc's avatar
SoleneDc committed
1207
      "Epoch 27/100\n",
SoleneDc's avatar
SoleneDc committed
1208
      "902/902 [==============================] - 0s 433us/step - loss: 0.4025 - acc: 0.8525 - val_loss: 0.6134 - val_acc: 0.8165\n",
SoleneDc's avatar
SoleneDc committed
1209
      "Epoch 28/100\n",
SoleneDc's avatar
SoleneDc committed
1210
      "902/902 [==============================] - 0s 334us/step - loss: 0.4114 - acc: 0.8570 - val_loss: 0.6684 - val_acc: 0.8062\n",
SoleneDc's avatar
SoleneDc committed
1211
      "Epoch 29/100\n",
SoleneDc's avatar
SoleneDc committed
1212
      "902/902 [==============================] - 0s 294us/step - loss: 0.4040 - acc: 0.8470 - val_loss: 0.5682 - val_acc: 0.8320\n",
SoleneDc's avatar
SoleneDc committed
1213
      "Epoch 30/100\n",
SoleneDc's avatar
SoleneDc committed
1214
      "902/902 [==============================] - 0s 333us/step - loss: 0.3872 - acc: 0.8614 - val_loss: 0.5725 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1215
      "Epoch 31/100\n",
SoleneDc's avatar
SoleneDc committed
1216
      "902/902 [==============================] - 0s 335us/step - loss: 0.3965 - acc: 0.8470 - val_loss: 0.6118 - val_acc: 0.8114\n",
SoleneDc's avatar
SoleneDc committed
1217
      "Epoch 32/100\n",
SoleneDc's avatar
SoleneDc committed
1218
      "902/902 [==============================] - 0s 310us/step - loss: 0.3892 - acc: 0.8581 - val_loss: 0.6597 - val_acc: 0.8140\n",
SoleneDc's avatar
SoleneDc committed
1219
      "Epoch 33/100\n",
SoleneDc's avatar
SoleneDc committed
1220
      "902/902 [==============================] - ETA: 0s - loss: 0.3921 - acc: 0.846 - 0s 304us/step - loss: 0.4135 - acc: 0.8426 - val_loss: 0.6365 - val_acc: 0.8165\n",
SoleneDc's avatar
SoleneDc committed
1221
      "Epoch 34/100\n",
SoleneDc's avatar
SoleneDc committed
1222
      "902/902 [==============================] - 0s 296us/step - loss: 0.3994 - acc: 0.8548 - val_loss: 0.6047 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1223
      "Epoch 35/100\n",
SoleneDc's avatar
SoleneDc committed
1224
      "902/902 [==============================] - 0s 338us/step - loss: 0.4080 - acc: 0.8570 - val_loss: 0.5755 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1225
      "Epoch 36/100\n",
SoleneDc's avatar
SoleneDc committed
1226
      "902/902 [==============================] - 0s 308us/step - loss: 0.4013 - acc: 0.8592 - val_loss: 0.5892 - val_acc: 0.8165\n",
SoleneDc's avatar
SoleneDc committed
1227
      "Epoch 37/100\n",
SoleneDc's avatar
SoleneDc committed
1228
      "902/902 [==============================] - 0s 336us/step - loss: 0.3899 - acc: 0.8559 - val_loss: 0.5975 - val_acc: 0.8191\n",
SoleneDc's avatar
SoleneDc committed
1229
      "Epoch 38/100\n",
SoleneDc's avatar
SoleneDc committed
1230
      "902/902 [==============================] - 0s 365us/step - loss: 0.3995 - acc: 0.8525 - val_loss: 0.5779 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1231
      "Epoch 39/100\n",
SoleneDc's avatar
SoleneDc committed
1232
      "902/902 [==============================] - 0s 291us/step - loss: 0.3794 - acc: 0.8636 - val_loss: 0.5914 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1233
      "Epoch 40/100\n",
SoleneDc's avatar
SoleneDc committed
1234
      "902/902 [==============================] - 0s 335us/step - loss: 0.3879 - acc: 0.8570 - val_loss: 0.6057 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1235
      "Epoch 41/100\n",
SoleneDc's avatar
SoleneDc committed
1236
      "902/902 [==============================] - 0s 290us/step - loss: 0.3924 - acc: 0.8525 - val_loss: 0.5841 - val_acc: 0.8165\n",
SoleneDc's avatar
SoleneDc committed
1237
      "Epoch 42/100\n",
SoleneDc's avatar
SoleneDc committed
1238
      "902/902 [==============================] - 0s 290us/step - loss: 0.3927 - acc: 0.8525 - val_loss: 0.6327 - val_acc: 0.8140\n",
SoleneDc's avatar
SoleneDc committed
1239
      "Epoch 43/100\n",
SoleneDc's avatar
SoleneDc committed
1240
      "902/902 [==============================] - 0s 324us/step - loss: 0.3973 - acc: 0.8492 - val_loss: 0.6064 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1241
      "Epoch 44/100\n",
SoleneDc's avatar
SoleneDc committed
1242
      "902/902 [==============================] - 0s 287us/step - loss: 0.3802 - acc: 0.8603 - val_loss: 0.6004 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1243
      "Epoch 45/100\n",
SoleneDc's avatar
SoleneDc committed
1244
      "902/902 [==============================] - 0s 299us/step - loss: 0.3736 - acc: 0.8570 - val_loss: 0.6174 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1245
      "Epoch 46/100\n",
SoleneDc's avatar
SoleneDc committed
1246
      "902/902 [==============================] - 0s 325us/step - loss: 0.3846 - acc: 0.8525 - val_loss: 0.6262 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1247
      "Epoch 47/100\n",
SoleneDc's avatar
SoleneDc committed
1248
      "902/902 [==============================] - 0s 287us/step - loss: 0.3857 - acc: 0.8548 - val_loss: 0.6079 - val_acc: 0.8191\n",
SoleneDc's avatar
SoleneDc committed
1249
      "Epoch 48/100\n",
SoleneDc's avatar
SoleneDc committed
1250
      "902/902 [==============================] - 0s 325us/step - loss: 0.3786 - acc: 0.8614 - val_loss: 0.5999 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1251
      "Epoch 49/100\n",
SoleneDc's avatar
SoleneDc committed
1252
      "902/902 [==============================] - 0s 292us/step - loss: 0.4185 - acc: 0.8459 - val_loss: 0.6327 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1253
      "Epoch 50/100\n",
SoleneDc's avatar
SoleneDc committed
1254
      "902/902 [==============================] - 0s 288us/step - loss: 0.3941 - acc: 0.8514 - val_loss: 0.6325 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1255
      "Epoch 51/100\n",
SoleneDc's avatar
SoleneDc committed
1256
      "902/902 [==============================] - 0s 336us/step - loss: 0.4053 - acc: 0.8459 - val_loss: 0.6480 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1257
      "Epoch 52/100\n",
SoleneDc's avatar
SoleneDc committed
1258
      "902/902 [==============================] - 0s 297us/step - loss: 0.3975 - acc: 0.8537 - val_loss: 0.6566 - val_acc: 0.8165\n",
SoleneDc's avatar
SoleneDc committed
1259
      "Epoch 53/100\n",
SoleneDc's avatar
SoleneDc committed
1260
      "902/902 [==============================] - 0s 307us/step - loss: 0.4011 - acc: 0.8492 - val_loss: 0.6257 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1261
      "Epoch 54/100\n",
SoleneDc's avatar
SoleneDc committed
1262
      "902/902 [==============================] - 0s 296us/step - loss: 0.4121 - acc: 0.8459 - val_loss: 0.6109 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1263
      "Epoch 55/100\n",
SoleneDc's avatar
SoleneDc committed
1264
      "902/902 [==============================] - ETA: 0s - loss: 0.3527 - acc: 0.877 - 0s 293us/step - loss: 0.3725 - acc: 0.8636 - val_loss: 0.6201 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1265
      "Epoch 56/100\n",
SoleneDc's avatar
SoleneDc committed
1266
      "902/902 [==============================] - 0s 339us/step - loss: 0.3659 - acc: 0.8647 - val_loss: 0.6236 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1267
      "Epoch 57/100\n",
SoleneDc's avatar
SoleneDc committed
1268
      "902/902 [==============================] - 0s 292us/step - loss: 0.3791 - acc: 0.8614 - val_loss: 0.6148 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1269
      "Epoch 58/100\n",
SoleneDc's avatar
SoleneDc committed
1270
      "902/902 [==============================] - 0s 299us/step - loss: 0.4070 - acc: 0.8537 - val_loss: 0.6759 - val_acc: 0.8088\n",
SoleneDc's avatar
SoleneDc committed
1271
      "Epoch 59/100\n",
SoleneDc's avatar
SoleneDc committed
1272 1273
      "902/902 [==============================] - 0s 318us/step - loss: 0.3712 - acc: 0.8659 - val_loss: 0.6309 - val_acc: 0.8295\n",
      "Epoch 60/100\n"
SoleneDc's avatar
SoleneDc committed
1274 1275 1276 1277 1278 1279
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
SoleneDc's avatar
SoleneDc committed
1280
      "902/902 [==============================] - 0s 291us/step - loss: 0.3834 - acc: 0.8570 - val_loss: 0.6710 - val_acc: 0.8191\n",
1281
      "Epoch 61/100\n",
SoleneDc's avatar
SoleneDc committed
1282
      "902/902 [==============================] - 0s 340us/step - loss: 0.3732 - acc: 0.8670 - val_loss: 0.6471 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1283
      "Epoch 62/100\n",
SoleneDc's avatar
SoleneDc committed
1284
      "902/902 [==============================] - 0s 296us/step - loss: 0.3633 - acc: 0.8603 - val_loss: 0.6557 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1285
      "Epoch 63/100\n",
SoleneDc's avatar
SoleneDc committed
1286
      "902/902 [==============================] - 0s 296us/step - loss: 0.3578 - acc: 0.8714 - val_loss: 0.6599 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1287
      "Epoch 64/100\n",
SoleneDc's avatar
SoleneDc committed
1288
      "902/902 [==============================] - 0s 333us/step - loss: 0.4056 - acc: 0.8415 - val_loss: 0.6269 - val_acc: 0.8191\n",
SoleneDc's avatar
SoleneDc committed
1289
      "Epoch 65/100\n",
SoleneDc's avatar
SoleneDc committed
1290
      "902/902 [==============================] - 0s 290us/step - loss: 0.3733 - acc: 0.8581 - val_loss: 0.5726 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1291
      "Epoch 66/100\n",
SoleneDc's avatar
SoleneDc committed
1292
      "902/902 [==============================] - 0s 334us/step - loss: 0.3715 - acc: 0.8581 - val_loss: 0.5643 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1293
      "Epoch 67/100\n",
SoleneDc's avatar
SoleneDc committed
1294
      "902/902 [==============================] - 0s 292us/step - loss: 0.3695 - acc: 0.8670 - val_loss: 0.5855 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1295
      "Epoch 68/100\n",
SoleneDc's avatar
SoleneDc committed
1296
      "902/902 [==============================] - 0s 288us/step - loss: 0.3770 - acc: 0.8647 - val_loss: 0.5929 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1297
      "Epoch 69/100\n",
SoleneDc's avatar
SoleneDc committed
1298
      "902/902 [==============================] - 0s 341us/step - loss: 0.3787 - acc: 0.8603 - val_loss: 0.6451 - val_acc: 0.8062\n",
SoleneDc's avatar
SoleneDc committed
1299
      "Epoch 70/100\n",
SoleneDc's avatar
SoleneDc committed
1300
      "902/902 [==============================] - 0s 309us/step - loss: 0.3715 - acc: 0.8614 - val_loss: 0.5979 - val_acc: 0.8191\n",
SoleneDc's avatar
SoleneDc committed
1301
      "Epoch 71/100\n",
SoleneDc's avatar
SoleneDc committed
1302
      "902/902 [==============================] - 0s 308us/step - loss: 0.3835 - acc: 0.8592 - val_loss: 0.6376 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1303
      "Epoch 72/100\n",
SoleneDc's avatar
SoleneDc committed
1304
      "902/902 [==============================] - 0s 307us/step - loss: 0.3602 - acc: 0.8703 - val_loss: 0.6588 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1305
      "Epoch 73/100\n",
SoleneDc's avatar
SoleneDc committed
1306
      "902/902 [==============================] - 0s 297us/step - loss: 0.3574 - acc: 0.8747 - val_loss: 0.6519 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1307
      "Epoch 74/100\n",
SoleneDc's avatar
SoleneDc committed
1308
      "902/902 [==============================] - 0s 339us/step - loss: 0.3545 - acc: 0.8736 - val_loss: 0.6611 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1309
      "Epoch 75/100\n",
SoleneDc's avatar
SoleneDc committed
1310
      "902/902 [==============================] - 0s 299us/step - loss: 0.3905 - acc: 0.8481 - val_loss: 0.5961 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1311
      "Epoch 76/100\n",
SoleneDc's avatar
SoleneDc committed
1312
      "902/902 [==============================] - 0s 305us/step - loss: 0.3737 - acc: 0.8592 - val_loss: 0.5874 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1313
      "Epoch 77/100\n",
SoleneDc's avatar
SoleneDc committed
1314
      "902/902 [==============================] - 0s 348us/step - loss: 0.3699 - acc: 0.8636 - val_loss: 0.6199 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1315
      "Epoch 78/100\n",
SoleneDc's avatar
SoleneDc committed
1316
      "902/902 [==============================] - 0s 298us/step - loss: 0.3714 - acc: 0.8692 - val_loss: 0.6230 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1317
      "Epoch 79/100\n",
SoleneDc's avatar
SoleneDc committed
1318
      "902/902 [==============================] - 0s 329us/step - loss: 0.3847 - acc: 0.8559 - val_loss: 0.5904 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1319
      "Epoch 80/100\n",
SoleneDc's avatar
SoleneDc committed
1320
      "902/902 [==============================] - 0s 295us/step - loss: 0.3993 - acc: 0.8514 - val_loss: 0.5974 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1321
      "Epoch 81/100\n",
SoleneDc's avatar
SoleneDc committed
1322
      "902/902 [==============================] - 0s 293us/step - loss: 0.3754 - acc: 0.8681 - val_loss: 0.6035 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1323
      "Epoch 82/100\n",
SoleneDc's avatar
SoleneDc committed
1324
      "902/902 [==============================] - 0s 333us/step - loss: 0.3669 - acc: 0.8647 - val_loss: 0.6539 - val_acc: 0.8217\n",
SoleneDc's avatar
SoleneDc committed
1325
      "Epoch 83/100\n",
SoleneDc's avatar
SoleneDc committed
1326
      "902/902 [==============================] - 0s 296us/step - loss: 0.3531 - acc: 0.8659 - val_loss: 0.6461 - val_acc: 0.8320\n",
SoleneDc's avatar
SoleneDc committed
1327
      "Epoch 84/100\n",
SoleneDc's avatar
SoleneDc committed
1328
      "902/902 [==============================] - 0s 310us/step - loss: 0.3847 - acc: 0.8659 - val_loss: 0.6137 - val_acc: 0.8398\n",
SoleneDc's avatar
SoleneDc committed
1329
      "Epoch 85/100\n",
SoleneDc's avatar
SoleneDc committed
1330
      "902/902 [==============================] - 0s 323us/step - loss: 0.3688 - acc: 0.8647 - val_loss: 0.6087 - val_acc: 0.8398\n",
SoleneDc's avatar
SoleneDc committed
1331
      "Epoch 86/100\n",
SoleneDc's avatar
SoleneDc committed
1332
      "902/902 [==============================] - 0s 294us/step - loss: 0.3472 - acc: 0.8714 - val_loss: 0.5913 - val_acc: 0.8398\n",
SoleneDc's avatar
SoleneDc committed
1333
      "Epoch 87/100\n",
SoleneDc's avatar
SoleneDc committed
1334
      "902/902 [==============================] - 0s 325us/step - loss: 0.3611 - acc: 0.8625 - val_loss: 0.5863 - val_acc: 0.8372\n",
SoleneDc's avatar
SoleneDc committed
1335
      "Epoch 88/100\n",
SoleneDc's avatar
SoleneDc committed
1336
      "902/902 [==============================] - 0s 302us/step - loss: 0.3639 - acc: 0.8636 - val_loss: 0.6222 - val_acc: 0.8346\n",
SoleneDc's avatar
SoleneDc committed
1337
      "Epoch 89/100\n",
SoleneDc's avatar
SoleneDc committed
1338
      "902/902 [==============================] - 0s 367us/step - loss: 0.3717 - acc: 0.8647 - val_loss: 0.6245 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1339
      "Epoch 90/100\n",
SoleneDc's avatar
SoleneDc committed
1340
      "902/902 [==============================] - 0s 548us/step - loss: 0.3592 - acc: 0.8647 - val_loss: 0.6098 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1341
      "Epoch 91/100\n",
SoleneDc's avatar
SoleneDc committed
1342
      "902/902 [==============================] - 0s 411us/step - loss: 0.3517 - acc: 0.8659 - val_loss: 0.6109 - val_acc: 0.8295\n",
SoleneDc's avatar
SoleneDc committed
1343
      "Epoch 92/100\n",
SoleneDc's avatar
SoleneDc committed
1344
      "902/902 [==============================] - 0s 352us/step - loss: 0.3744 - acc: 0.8603 - val_loss: 0.5732 - val_acc: 0.8269\n",
SoleneDc's avatar
SoleneDc committed
1345
      "Epoch 93/100\n",
SoleneDc's avatar
SoleneDc committed
1346
      "902/902 [==============================] - 0s 479us/step - loss: 0.3947 - acc: 0.8603 - val_loss: 0.6055 - val_acc: 0.8191\n",
SoleneDc's avatar
SoleneDc committed
1347
      "Epoch 94/100\n",
SoleneDc's avatar
SoleneDc committed
1348
      "902/902 [==============================] - 0s 358us/step - loss: 0.3682 - acc: 0.8647 - val_loss: 0.6374 - val_acc: 0.8243\n",
SoleneDc's avatar
SoleneDc committed
1349
      "Epoch 95/100\n",
SoleneDc's avatar
SoleneDc committed
1350
      "902/902 [==============================] - 0s 361us/step - loss: 0.3800 - acc: 0.8625 - val_loss: 0.5957 - val_acc: 0.8320\n",
SoleneDc's avatar
SoleneDc committed
1351
      "Epoch 96/100\n",
SoleneDc's avatar
SoleneDc committed
1352
      "902/902 [==============================] - 0s 331us/step - loss: 0.3808 - acc: 0.8636 - val_loss: 0.5638 - val_acc: 0.8346\n",
SoleneDc's avatar
SoleneDc committed
1353
      "Epoch 97/100\n",
SoleneDc's avatar
SoleneDc committed
1354
      "902/902 [==============================] - 0s 302us/step - loss: 0.3649 - acc: 0.8681 - val_loss: 0.5768 - val_acc: 0.8346\n",
SoleneDc's avatar
SoleneDc committed
1355
      "Epoch 98/100\n",
SoleneDc's avatar
SoleneDc committed
1356
      "902/902 [==============================] - 0s 356us/step - loss: 0.3696 - acc: 0.8592 - val_loss: 0.5707 - val_acc: 0.8320\n",
SoleneDc's avatar
SoleneDc committed
1357
      "Epoch 99/100\n",
SoleneDc's avatar
SoleneDc committed
1358
      "902/902 [==============================] - 0s 299us/step - loss: 0.3572 - acc: 0.8670 - val_loss: 0.5849 - val_acc: 0.8346\n",
SoleneDc's avatar
SoleneDc committed
1359
      "Epoch 100/100\n",
SoleneDc's avatar
SoleneDc committed
1360
      "902/902 [==============================] - 0s 461us/step - loss: 0.3545 - acc: 0.8769 - val_loss: 0.6003 - val_acc: 0.8320\n"
SoleneDc's avatar
SoleneDc committed
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373
     ]
    }
   ],
   "source": [
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "history = model.fit(X_train, Y_train, epochs=100, validation_split=0.3, batch_size=50 , verbose=1)"
   ]
  },
  {
   "cell_type": "code",
SoleneDc's avatar
SoleneDc committed
1374
   "execution_count": 112,
SoleneDc's avatar
SoleneDc committed
1375 1376
   "metadata": {},
   "outputs": [
1377 1378 1379 1380 1381
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
SoleneDc's avatar
SoleneDc committed
1382
      "       Année: 2007   //  Genre: 28  //  Données X_train: (1289, 1200, 3)\n",
1383 1384 1385
      "\n"
     ]
    },
SoleneDc's avatar
SoleneDc committed
1386 1387
    {
     "data": {
SoleneDc's avatar
SoleneDc committed
1388
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEWCAYAAABMoxE0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
SoleneDc's avatar
SoleneDc committed
1389
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
1390
       "<matplotlib.figure.Figure at 0x182393f400>"
SoleneDc's avatar
SoleneDc committed
1391 1392 1393 1394 1395 1396 1397
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
SoleneDc's avatar
SoleneDc committed
1398
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
SoleneDc's avatar
SoleneDc committed
1399
      "text/plain": [
SoleneDc's avatar
SoleneDc committed
1400
       "<matplotlib.figure.Figure at 0x183bd48f98>"
SoleneDc's avatar
SoleneDc committed
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "train, =plt.plot(history.history['acc'], label='Train set')\n",
    "val, =plt.plot(history.history['val_acc'], label='Validation set')\n",
1411
    "print('')\n",
SoleneDc's avatar
SoleneDc committed
1412
    "print(f\"       Année: {year}   //  Genre: {selected_genre}  //  Données X_train: {X_train.shape}\")\n",
1413
    "print('')\n",
SoleneDc's avatar
SoleneDc committed
1414
    "plt.title('model accuracy')\n",
1415
    "plt.ylabel(\"Accuracy\")\n",
SoleneDc's avatar
SoleneDc committed
1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459
    "plt.xlabel('Epochs')\n",
    "plt.legend(handles=[train, val])\n",
    "plt.show()\n",
    "train, =plt.plot(history.history['loss'], label='Train set')\n",
    "val, =plt.plot(history.history['val_loss'], label='Validation set')\n",
    "plt.title('model Loss')\n",
    "plt.ylabel('Cross entropy')\n",
    "plt.xlabel('Epochs')\n",
    "plt.legend(handles=[train, val])\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}