neural_nets_intro.ipynb

      "Epoch 146/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1994 - acc: 0.9600 - val_loss: 0.2480 - val_acc: 0.9400\n",
      "Epoch 147/300\n",
      "350/350 [==============================] - 0s 87us/step - loss: 0.1982 - acc: 0.9600 - val_loss: 0.2468 - val_acc: 0.9400\n",
      "Epoch 148/300\n",
      "350/350 [==============================] - 0s 91us/step - loss: 0.1973 - acc: 0.9600 - val_loss: 0.2456 - val_acc: 0.9400\n",
      "Epoch 149/300\n",
      "350/350 [==============================] - 0s 101us/step - loss: 0.1960 - acc: 0.9629 - val_loss: 0.2439 - val_acc: 0.9400\n",
      "Epoch 150/300\n",
      "350/350 [==============================] - 0s 96us/step - loss: 0.1949 - acc: 0.9600 - val_loss: 0.2429 - val_acc: 0.9400\n",
      "Epoch 151/300\n",
      "350/350 [==============================] - 0s 116us/step - loss: 0.1937 - acc: 0.9629 - val_loss: 0.2418 - val_acc: 0.9400\n",
      "Epoch 152/300\n",
      "350/350 [==============================] - 0s 97us/step - loss: 0.1925 - acc: 0.9600 - val_loss: 0.2403 - val_acc: 0.9333\n",
      "Epoch 153/300\n",
      "350/350 [==============================] - 0s 99us/step - loss: 0.1916 - acc: 0.9629 - val_loss: 0.2391 - val_acc: 0.9333\n",
      "Epoch 154/300\n",
      "350/350 [==============================] - 0s 94us/step - loss: 0.1904 - acc: 0.9629 - val_loss: 0.2376 - val_acc: 0.9333\n",
      "Epoch 155/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1895 - acc: 0.9629 - val_loss: 0.2364 - val_acc: 0.9333\n",
      "Epoch 156/300\n",
      "350/350 [==============================] - 0s 82us/step - loss: 0.1886 - acc: 0.9629 - val_loss: 0.2349 - val_acc: 0.9333\n",
      "Epoch 157/300\n",
      "350/350 [==============================] - 0s 92us/step - loss: 0.1873 - acc: 0.9629 - val_loss: 0.2343 - val_acc: 0.9333\n",
      "Epoch 158/300\n",
      "350/350 [==============================] - 0s 82us/step - loss: 0.1864 - acc: 0.9629 - val_loss: 0.2329 - val_acc: 0.9333\n",
      "Epoch 159/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1854 - acc: 0.9629 - val_loss: 0.2319 - val_acc: 0.9333\n",
      "Epoch 160/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1845 - acc: 0.9629 - val_loss: 0.2307 - val_acc: 0.9333\n",
      "Epoch 161/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1833 - acc: 0.9629 - val_loss: 0.2297 - val_acc: 0.9333\n",
      "Epoch 162/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1824 - acc: 0.9629 - val_loss: 0.2285 - val_acc: 0.9333\n",
      "Epoch 163/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1815 - acc: 0.9629 - val_loss: 0.2272 - val_acc: 0.9333\n",
      "Epoch 164/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1805 - acc: 0.9629 - val_loss: 0.2265 - val_acc: 0.9333\n",
      "Epoch 165/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1796 - acc: 0.9629 - val_loss: 0.2256 - val_acc: 0.9333\n",
      "Epoch 166/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1791 - acc: 0.9629 - val_loss: 0.2245 - val_acc: 0.9333\n",
      "Epoch 167/300\n",
      "350/350 [==============================] - 0s 110us/step - loss: 0.1781 - acc: 0.9629 - val_loss: 0.2234 - val_acc: 0.9333\n",
      "Epoch 168/300\n",
      "350/350 [==============================] - 0s 100us/step - loss: 0.1772 - acc: 0.9629 - val_loss: 0.2228 - val_acc: 0.9333\n",
      "Epoch 169/300\n",
      "350/350 [==============================] - 0s 101us/step - loss: 0.1761 - acc: 0.9629 - val_loss: 0.2214 - val_acc: 0.9333\n",
      "Epoch 170/300\n",
      "350/350 [==============================] - 0s 98us/step - loss: 0.1754 - acc: 0.9657 - val_loss: 0.2207 - val_acc: 0.9333\n",
      "Epoch 171/300\n",
      "350/350 [==============================] - 0s 97us/step - loss: 0.1745 - acc: 0.9629 - val_loss: 0.2198 - val_acc: 0.9333\n",
      "Epoch 172/300\n",
      "350/350 [==============================] - 0s 87us/step - loss: 0.1740 - acc: 0.9629 - val_loss: 0.2183 - val_acc: 0.9333\n",
      "Epoch 173/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1727 - acc: 0.9629 - val_loss: 0.2173 - val_acc: 0.9400\n",
      "Epoch 174/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1720 - acc: 0.9629 - val_loss: 0.2163 - val_acc: 0.9400\n",
      "Epoch 175/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1711 - acc: 0.9629 - val_loss: 0.2155 - val_acc: 0.9400\n",
      "Epoch 176/300\n",
      "350/350 [==============================] - 0s 88us/step - loss: 0.1702 - acc: 0.9629 - val_loss: 0.2143 - val_acc: 0.9400\n",
      "Epoch 177/300\n",
      "350/350 [==============================] - 0s 92us/step - loss: 0.1693 - acc: 0.9629 - val_loss: 0.2130 - val_acc: 0.9467\n",
      "Epoch 178/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1687 - acc: 0.9657 - val_loss: 0.2124 - val_acc: 0.9400\n",
      "Epoch 179/300\n",
      "350/350 [==============================] - 0s 77us/step - loss: 0.1677 - acc: 0.9629 - val_loss: 0.2111 - val_acc: 0.9533\n",
      "Epoch 180/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1671 - acc: 0.9629 - val_loss: 0.2098 - val_acc: 0.9533\n",
      "Epoch 181/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1664 - acc: 0.9629 - val_loss: 0.2093 - val_acc: 0.9533\n",
      "Epoch 182/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1658 - acc: 0.9629 - val_loss: 0.2082 - val_acc: 0.9533\n",
      "Epoch 183/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1650 - acc: 0.9629 - val_loss: 0.2071 - val_acc: 0.9533\n",
      "Epoch 184/300\n",
      "350/350 [==============================] - 0s 82us/step - loss: 0.1643 - acc: 0.9629 - val_loss: 0.2063 - val_acc: 0.9533\n",
      "Epoch 185/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1635 - acc: 0.9629 - val_loss: 0.2052 - val_acc: 0.9533\n",
      "Epoch 186/300\n",
      "350/350 [==============================] - 0s 92us/step - loss: 0.1629 - acc: 0.9629 - val_loss: 0.2044 - val_acc: 0.9533\n",
      "Epoch 187/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1621 - acc: 0.9629 - val_loss: 0.2037 - val_acc: 0.9533\n",
      "Epoch 188/300\n",
      "350/350 [==============================] - 0s 80us/step - loss: 0.1615 - acc: 0.9629 - val_loss: 0.2028 - val_acc: 0.9533\n",
      "Epoch 189/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1605 - acc: 0.9629 - val_loss: 0.2024 - val_acc: 0.9533\n",
      "Epoch 190/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1602 - acc: 0.9657 - val_loss: 0.2011 - val_acc: 0.9533\n",
      "Epoch 191/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1593 - acc: 0.9629 - val_loss: 0.2007 - val_acc: 0.9533\n",
      "Epoch 192/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1587 - acc: 0.9629 - val_loss: 0.2001 - val_acc: 0.9533\n",
      "Epoch 193/300\n",
      "350/350 [==============================] - 0s 91us/step - loss: 0.1579 - acc: 0.9657 - val_loss: 0.1995 - val_acc: 0.9533\n",
      "Epoch 194/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1574 - acc: 0.9629 - val_loss: 0.1989 - val_acc: 0.9533\n",
      "Epoch 195/300\n",
      "350/350 [==============================] - 0s 93us/step - loss: 0.1569 - acc: 0.9629 - val_loss: 0.1986 - val_acc: 0.9533\n",
      "Epoch 196/300\n",
      "350/350 [==============================] - 0s 80us/step - loss: 0.1561 - acc: 0.9657 - val_loss: 0.1979 - val_acc: 0.9533\n",
      "Epoch 197/300\n",
      "350/350 [==============================] - 0s 82us/step - loss: 0.1555 - acc: 0.9629 - val_loss: 0.1967 - val_acc: 0.9533\n",
      "Epoch 198/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1548 - acc: 0.9629 - val_loss: 0.1957 - val_acc: 0.9533\n",
      "Epoch 199/300\n",
      "350/350 [==============================] - 0s 76us/step - loss: 0.1543 - acc: 0.9629 - val_loss: 0.1950 - val_acc: 0.9533\n",
      "Epoch 200/300\n",
      "350/350 [==============================] - 0s 87us/step - loss: 0.1538 - acc: 0.9629 - val_loss: 0.1945 - val_acc: 0.9533\n",
      "Epoch 201/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1528 - acc: 0.9629 - val_loss: 0.1937 - val_acc: 0.9533\n",
      "Epoch 202/300\n",
      "350/350 [==============================] - 0s 75us/step - loss: 0.1522 - acc: 0.9629 - val_loss: 0.1926 - val_acc: 0.9533\n",
      "Epoch 203/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1519 - acc: 0.9629 - val_loss: 0.1927 - val_acc: 0.9533\n",
      "Epoch 204/300\n",
      "350/350 [==============================] - 0s 76us/step - loss: 0.1511 - acc: 0.9629 - val_loss: 0.1919 - val_acc: 0.9533\n",
      "Epoch 205/300\n",
      "350/350 [==============================] - 0s 76us/step - loss: 0.1504 - acc: 0.9657 - val_loss: 0.1907 - val_acc: 0.9533\n",
      "Epoch 206/300\n",
      "350/350 [==============================] - 0s 82us/step - loss: 0.1499 - acc: 0.9629 - val_loss: 0.1893 - val_acc: 0.9533\n",
      "Epoch 207/300\n",
      "350/350 [==============================] - 0s 91us/step - loss: 0.1496 - acc: 0.9657 - val_loss: 0.1886 - val_acc: 0.9533\n",
      "Epoch 208/300\n",
      "350/350 [==============================] - 0s 89us/step - loss: 0.1486 - acc: 0.9657 - val_loss: 0.1875 - val_acc: 0.9533\n",
      "Epoch 209/300\n",
      "350/350 [==============================] - 0s 93us/step - loss: 0.1481 - acc: 0.9657 - val_loss: 0.1866 - val_acc: 0.9533\n",
      "Epoch 210/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1475 - acc: 0.9657 - val_loss: 0.1858 - val_acc: 0.9533\n",
      "Epoch 211/300\n",
      "350/350 [==============================] - 0s 74us/step - loss: 0.1468 - acc: 0.9657 - val_loss: 0.1856 - val_acc: 0.9533\n",
      "Epoch 212/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1465 - acc: 0.9657 - val_loss: 0.1844 - val_acc: 0.9533\n",
      "Epoch 213/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1458 - acc: 0.9657 - val_loss: 0.1840 - val_acc: 0.9533\n",
      "Epoch 214/300\n",
      "350/350 [==============================] - 0s 75us/step - loss: 0.1452 - acc: 0.9657 - val_loss: 0.1838 - val_acc: 0.9533\n",
      "Epoch 215/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1451 - acc: 0.9629 - val_loss: 0.1833 - val_acc: 0.9533\n",
      "Epoch 216/300\n",
      "350/350 [==============================] - 0s 91us/step - loss: 0.1442 - acc: 0.9657 - val_loss: 0.1820 - val_acc: 0.9533\n",
      "Epoch 217/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1437 - acc: 0.9657 - val_loss: 0.1812 - val_acc: 0.9533\n",
      "Epoch 218/300\n",
      "350/350 [==============================] - 0s 93us/step - loss: 0.1432 - acc: 0.9657 - val_loss: 0.1812 - val_acc: 0.9533\n",
      "Epoch 219/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1426 - acc: 0.9657 - val_loss: 0.1808 - val_acc: 0.9533\n",
      "Epoch 220/300\n",
      "350/350 [==============================] - 0s 96us/step - loss: 0.1419 - acc: 0.9629 - val_loss: 0.1800 - val_acc: 0.9533\n",
      "Epoch 221/300\n",
      "350/350 [==============================] - 0s 91us/step - loss: 0.1417 - acc: 0.9657 - val_loss: 0.1796 - val_acc: 0.9533\n",
      "Epoch 222/300\n",
      "350/350 [==============================] - 0s 94us/step - loss: 0.1412 - acc: 0.9657 - val_loss: 0.1784 - val_acc: 0.9533\n",
      "Epoch 223/300\n",
      "350/350 [==============================] - 0s 105us/step - loss: 0.1404 - acc: 0.9657 - val_loss: 0.1780 - val_acc: 0.9533\n",
      "Epoch 224/300\n",
      "350/350 [==============================] - 0s 77us/step - loss: 0.1397 - acc: 0.9657 - val_loss: 0.1771 - val_acc: 0.9533\n",
      "Epoch 225/300\n",
      "350/350 [==============================] - 0s 88us/step - loss: 0.1395 - acc: 0.9657 - val_loss: 0.1764 - val_acc: 0.9533\n",
      "Epoch 226/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1389 - acc: 0.9657 - val_loss: 0.1765 - val_acc: 0.9533\n",
      "Epoch 227/300\n",
      "350/350 [==============================] - 0s 88us/step - loss: 0.1382 - acc: 0.9657 - val_loss: 0.1760 - val_acc: 0.9533\n",
      "Epoch 228/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1381 - acc: 0.9657 - val_loss: 0.1747 - val_acc: 0.9533\n",
      "Epoch 229/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1372 - acc: 0.9657 - val_loss: 0.1740 - val_acc: 0.9533\n",
      "Epoch 230/300\n",
      "350/350 [==============================] - 0s 89us/step - loss: 0.1367 - acc: 0.9657 - val_loss: 0.1737 - val_acc: 0.9533\n",
      "Epoch 231/300\n",
      "350/350 [==============================] - 0s 93us/step - loss: 0.1361 - acc: 0.9657 - val_loss: 0.1730 - val_acc: 0.9533\n",
      "Epoch 232/300\n",
      "350/350 [==============================] - 0s 93us/step - loss: 0.1358 - acc: 0.9657 - val_loss: 0.1723 - val_acc: 0.9533\n",
      "Epoch 233/300\n",
      "350/350 [==============================] - 0s 93us/step - loss: 0.1352 - acc: 0.9657 - val_loss: 0.1713 - val_acc: 0.9533\n",
      "Epoch 234/300\n",
      "350/350 [==============================] - 0s 92us/step - loss: 0.1347 - acc: 0.9657 - val_loss: 0.1705 - val_acc: 0.9600\n",
      "Epoch 235/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1342 - acc: 0.9657 - val_loss: 0.1703 - val_acc: 0.9600\n",
      "Epoch 236/300\n",
      "350/350 [==============================] - 0s 93us/step - loss: 0.1336 - acc: 0.9657 - val_loss: 0.1692 - val_acc: 0.9600\n",
      "Epoch 237/300\n",
      "350/350 [==============================] - 0s 89us/step - loss: 0.1331 - acc: 0.9657 - val_loss: 0.1689 - val_acc: 0.9600\n",
      "Epoch 238/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1327 - acc: 0.9657 - val_loss: 0.1687 - val_acc: 0.9600\n",
      "Epoch 239/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1321 - acc: 0.9657 - val_loss: 0.1679 - val_acc: 0.9600\n",
      "Epoch 240/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1316 - acc: 0.9657 - val_loss: 0.1670 - val_acc: 0.9600\n",
      "Epoch 241/300\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "350/350 [==============================] - 0s 86us/step - loss: 0.1311 - acc: 0.9657 - val_loss: 0.1665 - val_acc: 0.9600\n",
      "Epoch 242/300\n",
      "350/350 [==============================] - 0s 88us/step - loss: 0.1305 - acc: 0.9657 - val_loss: 0.1661 - val_acc: 0.9600\n",
      "Epoch 243/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1303 - acc: 0.9657 - val_loss: 0.1655 - val_acc: 0.9600\n",
      "Epoch 244/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1296 - acc: 0.9657 - val_loss: 0.1655 - val_acc: 0.9600\n",
      "Epoch 245/300\n",
      "350/350 [==============================] - 0s 76us/step - loss: 0.1294 - acc: 0.9657 - val_loss: 0.1653 - val_acc: 0.9600\n",
      "Epoch 246/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1289 - acc: 0.9686 - val_loss: 0.1642 - val_acc: 0.9600\n",
      "Epoch 247/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1286 - acc: 0.9657 - val_loss: 0.1639 - val_acc: 0.9600\n",
      "Epoch 248/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1281 - acc: 0.9657 - val_loss: 0.1634 - val_acc: 0.9600\n",
      "Epoch 249/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1279 - acc: 0.9657 - val_loss: 0.1629 - val_acc: 0.9600\n",
      "Epoch 250/300\n",
      "350/350 [==============================] - 0s 76us/step - loss: 0.1274 - acc: 0.9657 - val_loss: 0.1621 - val_acc: 0.9600\n",
      "Epoch 251/300\n",
      "350/350 [==============================] - 0s 87us/step - loss: 0.1270 - acc: 0.9657 - val_loss: 0.1615 - val_acc: 0.9600\n",
      "Epoch 252/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1264 - acc: 0.9657 - val_loss: 0.1615 - val_acc: 0.9600\n",
      "Epoch 253/300\n",
      "350/350 [==============================] - 0s 75us/step - loss: 0.1263 - acc: 0.9657 - val_loss: 0.1610 - val_acc: 0.9600\n",
      "Epoch 254/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1257 - acc: 0.9657 - val_loss: 0.1600 - val_acc: 0.9600\n",
      "Epoch 255/300\n",
      "350/350 [==============================] - 0s 82us/step - loss: 0.1253 - acc: 0.9657 - val_loss: 0.1598 - val_acc: 0.9600\n",
      "Epoch 256/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1251 - acc: 0.9657 - val_loss: 0.1590 - val_acc: 0.9600\n",
      "Epoch 257/300\n",
      "350/350 [==============================] - 0s 90us/step - loss: 0.1244 - acc: 0.9657 - val_loss: 0.1587 - val_acc: 0.9600\n",
      "Epoch 258/300\n",
      "350/350 [==============================] - 0s 76us/step - loss: 0.1243 - acc: 0.9657 - val_loss: 0.1586 - val_acc: 0.9600\n",
      "Epoch 259/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1238 - acc: 0.9657 - val_loss: 0.1581 - val_acc: 0.9600\n",
      "Epoch 260/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1233 - acc: 0.9657 - val_loss: 0.1575 - val_acc: 0.9600\n",
      "Epoch 261/300\n",
      "350/350 [==============================] - 0s 75us/step - loss: 0.1230 - acc: 0.9657 - val_loss: 0.1568 - val_acc: 0.9600\n",
      "Epoch 262/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1226 - acc: 0.9657 - val_loss: 0.1569 - val_acc: 0.9600\n",
      "Epoch 263/300\n",
      "350/350 [==============================] - 0s 77us/step - loss: 0.1223 - acc: 0.9657 - val_loss: 0.1555 - val_acc: 0.9600\n",
      "Epoch 264/300\n",
      "350/350 [==============================] - 0s 80us/step - loss: 0.1217 - acc: 0.9657 - val_loss: 0.1547 - val_acc: 0.9600\n",
      "Epoch 265/300\n",
      "350/350 [==============================] - 0s 82us/step - loss: 0.1215 - acc: 0.9657 - val_loss: 0.1551 - val_acc: 0.9600\n",
      "Epoch 266/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1211 - acc: 0.9657 - val_loss: 0.1548 - val_acc: 0.9600\n",
      "Epoch 267/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1206 - acc: 0.9657 - val_loss: 0.1540 - val_acc: 0.9600\n",
      "Epoch 268/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1206 - acc: 0.9657 - val_loss: 0.1533 - val_acc: 0.9600\n",
      "Epoch 269/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1202 - acc: 0.9686 - val_loss: 0.1534 - val_acc: 0.9600\n",
      "Epoch 270/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1195 - acc: 0.9657 - val_loss: 0.1528 - val_acc: 0.9600\n",
      "Epoch 271/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1197 - acc: 0.9657 - val_loss: 0.1524 - val_acc: 0.9600\n",
      "Epoch 272/300\n",
      "350/350 [==============================] - 0s 75us/step - loss: 0.1190 - acc: 0.9657 - val_loss: 0.1518 - val_acc: 0.9600\n",
      "Epoch 273/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1188 - acc: 0.9657 - val_loss: 0.1520 - val_acc: 0.9600\n",
      "Epoch 274/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1184 - acc: 0.9657 - val_loss: 0.1519 - val_acc: 0.9600\n",
      "Epoch 275/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1181 - acc: 0.9657 - val_loss: 0.1510 - val_acc: 0.9600\n",
      "Epoch 276/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1178 - acc: 0.9657 - val_loss: 0.1501 - val_acc: 0.9600\n",
      "Epoch 277/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1173 - acc: 0.9657 - val_loss: 0.1496 - val_acc: 0.9600\n",
      "Epoch 278/300\n",
      "350/350 [==============================] - 0s 74us/step - loss: 0.1169 - acc: 0.9657 - val_loss: 0.1495 - val_acc: 0.9600\n",
      "Epoch 279/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1168 - acc: 0.9657 - val_loss: 0.1489 - val_acc: 0.9600\n",
      "Epoch 280/300\n",
      "350/350 [==============================] - 0s 75us/step - loss: 0.1165 - acc: 0.9657 - val_loss: 0.1488 - val_acc: 0.9600\n",
      "Epoch 281/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1162 - acc: 0.9657 - val_loss: 0.1476 - val_acc: 0.9600\n",
      "Epoch 282/300\n",
      "350/350 [==============================] - 0s 87us/step - loss: 0.1158 - acc: 0.9657 - val_loss: 0.1474 - val_acc: 0.9600\n",
      "Epoch 283/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1155 - acc: 0.9657 - val_loss: 0.1473 - val_acc: 0.9600\n",
      "Epoch 284/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1153 - acc: 0.9657 - val_loss: 0.1474 - val_acc: 0.9600\n",
      "Epoch 285/300\n",
      "350/350 [==============================] - 0s 74us/step - loss: 0.1152 - acc: 0.9686 - val_loss: 0.1472 - val_acc: 0.9600\n",
      "Epoch 286/300\n",
      "350/350 [==============================] - 0s 84us/step - loss: 0.1146 - acc: 0.9657 - val_loss: 0.1465 - val_acc: 0.9600\n",
      "Epoch 287/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1145 - acc: 0.9657 - val_loss: 0.1459 - val_acc: 0.9600\n",
      "Epoch 288/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1142 - acc: 0.9657 - val_loss: 0.1457 - val_acc: 0.9600\n",
      "Epoch 289/300\n",
      "350/350 [==============================] - 0s 83us/step - loss: 0.1139 - acc: 0.9657 - val_loss: 0.1454 - val_acc: 0.9600\n",
      "Epoch 290/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1136 - acc: 0.9657 - val_loss: 0.1454 - val_acc: 0.9600\n",
      "Epoch 291/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1133 - acc: 0.9657 - val_loss: 0.1448 - val_acc: 0.9600\n",
      "Epoch 292/300\n",
      "350/350 [==============================] - 0s 87us/step - loss: 0.1129 - acc: 0.9657 - val_loss: 0.1443 - val_acc: 0.9600\n",
      "Epoch 293/300\n",
      "350/350 [==============================] - 0s 88us/step - loss: 0.1127 - acc: 0.9657 - val_loss: 0.1440 - val_acc: 0.9600\n",
      "Epoch 294/300\n",
      "350/350 [==============================] - 0s 79us/step - loss: 0.1124 - acc: 0.9657 - val_loss: 0.1439 - val_acc: 0.9600\n",
      "Epoch 295/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1121 - acc: 0.9657 - val_loss: 0.1436 - val_acc: 0.9600\n",
      "Epoch 296/300\n",
      "350/350 [==============================] - 0s 78us/step - loss: 0.1116 - acc: 0.9657 - val_loss: 0.1429 - val_acc: 0.9600\n",
      "Epoch 297/300\n",
      "350/350 [==============================] - 0s 81us/step - loss: 0.1118 - acc: 0.9657 - val_loss: 0.1420 - val_acc: 0.9600\n",
      "Epoch 298/300\n",
      "350/350 [==============================] - 0s 85us/step - loss: 0.1113 - acc: 0.9657 - val_loss: 0.1418 - val_acc: 0.9600\n",
      "Epoch 299/300\n",
      "350/350 [==============================] - 0s 86us/step - loss: 0.1110 - acc: 0.9657 - val_loss: 0.1414 - val_acc: 0.9600\n",
      "Epoch 300/300\n",
      "350/350 [==============================] - 0s 94us/step - loss: 0.1106 - acc: 0.9657 - val_loss: 0.1417 - val_acc: 0.9600\n"
     ]
    }
   ],
   "source": [
    "# Instantiating the model\n",
    "model = a_simple_NN()\n",
    "\n",
    "# Splitting the dataset into training (70%) and validation sets (30%)\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    features, labels, test_size=0.3)\n",
    "\n",
    "# Setting the number of passes through the entire training set\n",
    "num_epochs = 300\n",
    "\n",
    "# We can pass validation data while training\n",
    "model_run = model.fit(X_train, y_train, epochs=num_epochs,\n",
    "                      validation_data=(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The history has the following data:  dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Looking at the loss and accuracy on the training and validation sets during the training\n",
    "# This can be done by using Keras callback \"history\" which is applied by default\n",
    "history_model = model_run.history\n",
    "\n",
    "print(\"The history has the following data: \", history_model.keys())\n",
    "\n",
    "# Plotting the training and validation accuracy during the training\n",
    "plt.plot(np.arange(1, num_epochs+1), history_model[\"acc\"], \"blue\") ;\n",
    "\n",
    "plt.plot(np.arange(1, num_epochs+1), history_model[\"val_acc\"], \"red\") ;"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Here we dont't really see a big difference between the training and validation data because the function we are trying to fit is quiet simple and there is not too much noise. We will come back to these curves in a later example**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the example above we splitted our dataset into a 70-30 train-validation set. We know from previous chapters that to more robustly calculate accuracy we can use **K-fold crossvalidation**.\n",
    "This is even more important when we have small datasets and cannot afford to reserve a validation set!\n",
    "\n",
    "One way to do the cross validation here would be to write our own function to do this. However, we also know that **SciKit learn** provides several handy functions to evaluate and tune the models. So the question is:\n",
    "\n",
    "Can we somehow use these **Scikit learn** functions or ones we wrote ourselves for **Scikit learn** models to evaluate and tune our Keras models?\n",
    "\n",
    "The Answer is **YES !**\n",
    "\n",
    "We show how to do this in the following section."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using SciKit learn functions on Keras models\n",
    "\n",
    "Keras offers 2 wrappers which allow its Sequential models to be used with SciKit learn. \n",
    "\n",
    "There are: **KerasClassifier** and **KerasRegressor**.\n",
    "\n",
    "For more information:\n",
    "https://keras.io/scikit-learn-api/\n",
    "\n",
    "**Now lets see how this works!**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We wrap the Keras model we created above with KerasClassifier\n",
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from sklearn.model_selection import cross_val_score\n",
    "# Wrapping Keras model\n",
    "# NOTE: We pass verbose=0 to suppress the model output\n",
    "num_epochs = 400\n",
    "model_scikit = KerasClassifier(\n",
    "    build_fn=a_simple_NN, **{\"epochs\": num_epochs, \"verbose\": 0})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's reuse the function to visualize the decision boundary which we saw in chapter 2 with minimal change\n",
    "\n",
    "def list_flatten(list_of_list):\n",
    "    flattened_list = [i for j in list_of_list for i in j]\n",
    "    return flattened_list\n",
    "\n",
    "def plot_points(plt=plt, marker='o'):\n",
    "    colors = [[\"steelblue\", \"chocolate\"][i] for i in labels]\n",
    "    plt.scatter(features.iloc[:, 0], features.iloc[:, 1], color=colors, marker=marker);\n",
    "\n",
    "def train_and_plot_decision_surface(\n",
    "    name, classifier, features_2d, labels, preproc=None, plt=plt, marker='o', N=400\n",
    "):\n",
    "\n",
    "    features_2d = np.array(features_2d)\n",
    "    xmin, ymin = features_2d.min(axis=0)\n",
    "    xmax, ymax = features_2d.max(axis=0)\n",
    "\n",
    "    x = np.linspace(xmin, xmax, N)\n",
    "    y = np.linspace(ymin, ymax, N)\n",
    "    points = np.array(np.meshgrid(x, y)).T.reshape(-1, 2)\n",
    "\n",
    "    if preproc is not None:\n",
    "        points_for_classifier = preproc.fit_transform(points)\n",
    "        features_2d = preproc.fit_transform(features_2d)\n",
    "    else:\n",
    "        points_for_classifier = points\n",
    "\n",
    "    classifier.fit(features_2d, labels, verbose=0)\n",
    "    predicted = classifier.predict(features_2d)\n",
    "    \n",
    "    if name == \"Neural Net\":\n",
    "        predicted = list_flatten(predicted)\n",
    "    \n",
    "    \n",
    "    if preproc is not None:\n",
    "        name += \" (w/ preprocessing)\"\n",
    "    print(name + \":\\t\", sum(predicted == labels), \"/\", len(labels), \"correct\")\n",
    "    \n",
    "    if name == \"Neural Net\":\n",
    "        classes = np.array(list_flatten(classifier.predict(points_for_classifier)), dtype=bool)\n",
    "    else:\n",
    "        classes = np.array(classifier.predict(points_for_classifier), dtype=bool)\n",
    "    plt.plot(\n",
    "        points[~classes][:, 0],\n",
    "        points[~classes][:, 1],\n",
    "        \"o\",\n",
    "        color=\"steelblue\",\n",
    "        markersize=1,\n",
    "        alpha=0.01,\n",
    "    )\n",
    "    plt.plot(\n",
    "        points[classes][:, 0],\n",
    "        points[classes][:, 1],\n",
    "        \"o\",\n",
    "        color=\"chocolate\",\n",
    "        markersize=1,\n",
    "        alpha=0.04,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Neural Net:\t 484 / 500 correct\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "_, ax = plt.subplots(figsize=(6, 6))\n",
    "\n",
    "train_and_plot_decision_surface(\"Neural Net\", model_scikit, features, labels, plt=ax)\n",
    "plot_points(plt=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The acuracy on the  5  validation folds: [0.72 0.94 0.95 0.75 0.94]\n",
      "The Average acuracy on the  5  validation folds: 0.86\n"
     ]
    }
   ],
   "source": [
    "# Applying K-fold cross-validation\n",
    "# Here we pass the whole dataset, i.e. features and labels, instead of splitting it.\n",
    "num_folds = 5\n",
    "cross_validation = cross_val_score(\n",
    "    model_scikit, features, labels, cv=num_folds, verbose=0)\n",
    "\n",
    "print(\"The acuracy on the \", num_folds, \" validation folds:\", cross_validation)\n",
    "print(\"The Average acuracy on the \", num_folds, \" validation folds:\", np.mean(cross_validation))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### NOTE: The above code took quiet long even though we used only 5  CV folds and the neural network and data size are very small!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hyperparameter optimization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We know from chapter 6 that there are 2 types of parameters which need to be tuned for a machine learning model.\n",
    "* Normal model parameters which can be learned for e.g. by gradient-descent\n",
    "* Hyperparameters\n",
    "\n",
    "In the model which we created above we made some arbitrary choices like which optimizer we use, what is its learning rate, number of hidden units and so on ...\n",
    "\n",
    "Now that we have the keras model wrapped as a scikit model we can use the grid search functions we have seen in chapter 6."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import GridSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9580000002384186 {'epochs': 500}\n"
     ]
    }
   ],
   "source": [
    "HP_grid = dict(epochs=[300, 500, 1000])\n",
    "search = GridSearchCV(estimator=model_scikit, param_grid=HP_grid)\n",
    "search.fit(features, labels)\n",
    "print(search.best_score_, search.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/tarunchadha/anaconda3/envs/mlw-2/lib/python3.6/site-packages/sklearn/model_selection/_search.py:841: DeprecationWarning: The default of the `iid` parameter will change from True to False in version 0.22 and will be removed in 0.24. This will change numeric results when test-set sizes are unequal.\n",
      "  DeprecationWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8299999971389771 {'batch_size': 30, 'epochs': 30}\n"
     ]
    }
   ],
   "source": [
    "HP_grid = {'epochs' : [10, 15, 30], \n",
    "           'batch_size' : [10, 20, 30] }\n",
    "search = GridSearchCV(estimator=model_scikit, param_grid=HP_grid)\n",
    "search.fit(features, labels)\n",
    "print(search.best_score_, search.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# A more general model for further Hyperparameter optimization\n",
    "from keras import optimizers\n",
    "\n",
    "def a_simple_NN(activation='relu', num_hidden_neurons=[4, 4], learning_rate=0.01):\n",
    "\n",
    "    model = Sequential()\n",
    "\n",
    "    model.add(Dense(num_hidden_neurons[0],\n",
    "                    input_shape=(2,), activation=activation))\n",
    "\n",
    "    model.add(Dense(num_hidden_neurons[1], activation=activation))\n",
    "\n",
    "    model.add(Dense(1, activation=\"sigmoid\"))\n",
    "\n",
    "    model.compile(loss=\"binary_crossentropy\", optimizer=optimizers.rmsprop(\n",
    "        lr=learning_rate), metrics=[\"accuracy\"])\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Exercise: \n",
    "* Look at the model above and choose a couple of hyperparameters to optimize. \n",
    "* What function from SciKit learn other than GridSearchCV can we use for hyperparameter optimization? Use it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Code here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Exercise: Create a neural network to classify the 2d points example from chapter 2 learned \n",
    "(Optional: As you create the model read a bit on the different keras commands we have used)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "circle = pd.read_csv(\"2d_points.csv\")\n",
    "# Using x and y coordinates as featues\n",
    "features = circle.iloc[:, :-1]\n",
    "# Convert boolean to integer values (True->1 and False->0)\n",
    "labels = circle.iloc[:, -1].astype(int)\n",
    "\n",
    "colors = [[\"steelblue\", \"chocolate\"][i] for i in circle[\"label\"]]\n",
    "plt.figure(figsize=(5, 5))\n",
    "plt.xlim([-2, 2])\n",
    "plt.ylim([-2, 2])\n",
    "\n",
    "plt.scatter(features[\"x\"], features[\"y\"], color=colors, marker=\"o\");\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Insert Code here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The examples above are not the ideal use problems one should use neural networks for. They are too simple and can be easily solved by classical machine learning algorithms. Below we show examples which are the more common applications of Neural Networks."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Handwritten Digits Classification\n",
    "### MNIST Dataset\n",
    "\n",
    "MNIST datasets is a very common dataset used in machine learning. It is widely used to train and validate models.\n",
    "\n",
    "\n",
    ">The MNIST database of handwritten digits, available from this page, has a training set of 60,000 examples, and a >test set of 10,000 examples. It is a subset of a larger set available from NIST. The digits have been size->normalized and centered in a fixed-size image.\n",
    ">It is a good database for people who want to try learning techniques and pattern recognition methods on real-world >data while spending minimal efforts on preprocessing and formatting.\n",
    ">source: http://yann.lecun.com/exdb/mnist/\n",
    "\n",
    "The problem we want to solve using this dataset is: multi-class classification\n",
    "This dataset consists of images of handwritten digits between 0-9 and their corresponsing labels. We want to train a neural network which is able to predict the correct digit on the image. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Loading the dataset in keras\n",
    "# Later you can explore and play with other datasets with come with Keras\n",
    "from keras.datasets import mnist\n",
    "\n",
    "# Loading the train and test data\n",
    "\n",
    "(X_train, y_train), (X_test, y_test) = mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 28, 28)\n"
     ]
    }
   ],
   "source": [
    "# Looking at the dataset\n",
    "print(X_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "This digit is:  0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# We can see that the training set consists of 60,000 images of size 28x28 pixels\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "i=np.random.randint(0,X_train.shape[0])\n",
    "plt.imshow(X_train[i], cmap=\"gray_r\") ;\n",
    "print(\"This digit is: \" , y_train[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 255\n"
     ]
    }
   ],
   "source": [
    "# Look at the data values for a couple of images\n",
    "print(X_train[0].min(), X_train[1].max())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data consists of values between 0-255 representing the **grayscale level**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000,)\n"
     ]
    }
   ],
   "source": [
    "# The labels are the digit on the image\n",
    "print(y_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scaling the data\n",
    "# It is important to normalize the input data to (0-1) before providing it to a neural net\n",
    "# We could use the previously introduced function from SciKit learn. However, here it is sufficient to\n",
    "# just divide the input data by 255\n",
    "X_train_norm = X_train/255.\n",
    "X_test_norm = X_test/255.\n",
    "\n",
    "# Also we need to reshape the input data such that each sample is a vector and not a 2D matrix\n",
    "X_train_prep = X_train_norm.reshape(X_train_norm.shape[0],28*28)\n",
    "X_test_prep = X_test_norm.reshape(X_test_norm.shape[0],28*28)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**IMPORTANT: One-Hot encoding**\n",
    "\n",
    "**TODO: Better frame the explaination**\n",
    "\n",
    "In such problems the labels are provided as something called **One-hot encodings**. What this does is to convert a categorical label to a vector.\n",
    "\n",
    "For the MNIST problem where we have **10 categories** one-hot encoding will create a vector of length 10 for each of the labels. All the entries of this vector will be zero **except** for the index which is equal to the integer value of the label.\n",
    "\n",
    "For example:\n",
    "if label is 4. The one-hot vector will look like **[0 0 0 0 1 0 0 0 0 0]**\n",
    "\n",
    "Fortunately, we don't have to code this ourselves because Keras has a built-in function for this."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 10)\n"
     ]
    }
   ],
   "source": [
    "from keras.utils.np_utils import to_categorical\n",
    "\n",
    "y_train_onehot = to_categorical(y_train, num_classes=10)\n",
    "y_test_onehot = to_categorical(y_test, num_classes=10)\n",
    "\n",
    "print(y_train_onehot.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 60000 samples, validate on 10000 samples\n",
      "Epoch 1/20\n",
      "60000/60000 [==============================] - 4s 60us/step - loss: 0.5733 - acc: 0.8465 - val_loss: 0.3185 - val_acc: 0.9084\n",
      "Epoch 2/20\n",
      "60000/60000 [==============================] - 1s 22us/step - loss: 0.2526 - acc: 0.9269 - val_loss: 0.2461 - val_acc: 0.9244\n",
      "Epoch 3/20\n",
      "60000/60000 [==============================] - 1s 23us/step - loss: 0.1955 - acc: 0.9427 - val_loss: 0.1884 - val_acc: 0.9423\n",
      "Epoch 4/20\n",
      "60000/60000 [==============================] - 1s 21us/step - loss: 0.1595 - acc: 0.9540 - val_loss: 0.1494 - val_acc: 0.9562\n",
      "Epoch 5/20\n",
      "60000/60000 [==============================] - 1s 23us/step - loss: 0.1353 - acc: 0.9601 - val_loss: 0.1602 - val_acc: 0.9502\n",
      "Epoch 6/20\n",
      "60000/60000 [==============================] - 1s 22us/step - loss: 0.1172 - acc: 0.9654 - val_loss: 0.1331 - val_acc: 0.9623\n",
      "Epoch 7/20\n",
      "60000/60000 [==============================] - 1s 23us/step - loss: 0.1040 - acc: 0.9692 - val_loss: 0.1227 - val_acc: 0.9629\n",
      "Epoch 8/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0932 - acc: 0.9720 - val_loss: 0.1422 - val_acc: 0.9569\n",
      "Epoch 9/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0839 - acc: 0.9752 - val_loss: 0.1090 - val_acc: 0.9668\n",
      "Epoch 10/20\n",
      "60000/60000 [==============================] - 1s 22us/step - loss: 0.0771 - acc: 0.9772 - val_loss: 0.1327 - val_acc: 0.9588\n",
      "Epoch 11/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0701 - acc: 0.9794 - val_loss: 0.1291 - val_acc: 0.9594\n",
      "Epoch 12/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0643 - acc: 0.9804 - val_loss: 0.1032 - val_acc: 0.9716\n",
      "Epoch 13/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0592 - acc: 0.9820 - val_loss: 0.1186 - val_acc: 0.9633\n",
      "Epoch 14/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0547 - acc: 0.9830 - val_loss: 0.0999 - val_acc: 0.9697\n",
      "Epoch 15/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0501 - acc: 0.9852 - val_loss: 0.1119 - val_acc: 0.9658\n",
      "Epoch 16/20\n",
      "60000/60000 [==============================] - 1s 23us/step - loss: 0.0461 - acc: 0.9862 - val_loss: 0.0948 - val_acc: 0.9712\n",
      "Epoch 17/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0431 - acc: 0.9875 - val_loss: 0.1183 - val_acc: 0.9642\n",
      "Epoch 18/20\n",
      "60000/60000 [==============================] - 1s 25us/step - loss: 0.0395 - acc: 0.9881 - val_loss: 0.1122 - val_acc: 0.9671\n",
      "Epoch 19/20\n",
      "60000/60000 [==============================] - 1s 23us/step - loss: 0.0360 - acc: 0.9892 - val_loss: 0.0943 - val_acc: 0.9724\n",
      "Epoch 20/20\n",
      "60000/60000 [==============================] - 1s 24us/step - loss: 0.0337 - acc: 0.9898 - val_loss: 0.1361 - val_acc: 0.9597\n"
     ]
    }
   ],
   "source": [
    "# Building the keras model\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "\n",
    "def mnist_model():\n",
    "    model = Sequential()\n",
    "\n",
    "    model.add(Dense(64, input_shape=(28*28,), activation=\"relu\"))\n",
    "\n",
    "    model.add(Dense(64, activation=\"relu\"))\n",
    "\n",
    "    model.add(Dense(10, activation=\"softmax\"))\n",
    "\n",
    "    model.compile(loss=\"categorical_crossentropy\",\n",
    "                  optimizer=\"rmsprop\", metrics=[\"accuracy\"])\n",
    "    return model\n",
    "\n",
    "model = mnist_model()\n",
    "\n",
    "model_run = model.fit(X_train_prep, y_train_onehot, epochs=20,\n",
    "                      batch_size=512)"
   ]