Redes Neurais/Redes Neurais com Keras.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c24be51a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importando as bibliotecas a serem utilizadas\n",
    "\n",
    "from sklearn import datasets\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import ConfusionMatrixDisplay\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.utils import np_utils\n",
    "from keras.utils.vis_utils import plot_model\n",
    "\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d562b48c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importando a base de dados que será utilizada\n",
    "\n",
    "base = datasets.load_iris()\n",
    "previsores = base.data\n",
    "classe = base.target\n",
    "classe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "631c7b8d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Transformação da classe para o formato \"dummy\", pois será feita uma rede neural com 3 neurônios na camada de saída\n",
    "\n",
    "classe_dummy = np_utils.to_categorical(classe)\n",
    "classe_dummy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1b7e8d03",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_treinamento, X_teste, Y_treinamento, Y_teste = train_test_split(previsores, classe_dummy, test_size = 0.3, random_state = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "09cbe95e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Criação da estrutura da rede neural com a classe Sequential (Com essa classe as camadas são adicionadas de forma sequencial)\n",
    "modelo = Sequential()\n",
    "\n",
    "# Primeira camada oculta, 5 neuronios, 4 neuronios de entrada (Funçã dense faz com que todos os neurônios conectem com a camada seguinte)\n",
    "modelo.add(Dense(units=5, input_dim=4))\n",
    "\n",
    "# Segunda camada oculta\n",
    "modelo.add(Dense(units = 4))\n",
    "\n",
    "# Terceira camada oculta\n",
    "# Função softmax porque temos um problema de classificação com mais de duas classes\n",
    "# (é gerada uma probabilidade em cada neurônio)\n",
    "modelo.add(Dense(units = 3, activation = 'softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f9631760",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense (Dense)               (None, 5)                 25        \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 4)                 24        \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 3)                 15        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 64\n",
      "Trainable params: 64\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "modelo.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cf7f30d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Configuração dos parâmetros da rede neural (adam = algoritmo para atualizar os pesos e loss = cálculo do erro)\n",
    "modelo.compile(optimizer = 'adam', loss = 'categorical_crossentropy', metrics = ['accuracy'])\n",
    "\n",
    "# Treinamento\n",
    "modelo.fit(X_treinamento, Y_treinamento, epochs = 1000, validation_data = (X_teste, Y_teste))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fbded2ec",
   "metadata": {},
   "outputs": [],
   "source": [
    "previsoes = modelo.predict(X_teste)\n",
    "previsoes = (previsoes > 0.5)\n",
    "previsoes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "06eb91ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "Y_teste_matrix = [np.argmax(t) for t in Y_teste]\n",
    "Y_previsao_matrix = [np.argmax(t) for t in previsoes]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "9e70ff39",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[16,  0,  0],\n",
       "       [ 0, 17,  1],\n",
       "       [ 0,  0, 11]], dtype=int64)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Gerando a matriz de confusão\n",
    "confusao = confusion_matrix(Y_teste_matrix, Y_previsao_matrix)\n",
    "confusao"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "2b88abf9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAe0AAAGsCAYAAADwsP8yAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAySUlEQVR4nO3de3gUZZr38V+HhCRAOhAggUDCQZHzGQQUMYwKRkHQVWFBRQUVOYlxEGcRwQNE3FmMwADK7ALjCIPvKgdPaFYFREVNAFFgQDRAFGNAkEBCSNJd7x9Ij21A06nuru7i+7muui7qqa6qu1PAnft5nqpyGIZhCAAAhLwIqwMAAABVQ9IGACBMkLQBAAgTJG0AAMIESRsAgDBB0gYAIEyQtAEACBMkbQAAwkSk1QFUldvt1qFDhxQXFyeHw2F1OAAAHxmGoRMnTig5OVkREYGrGUtLS1VWVmb6ODVr1lRMTIwfIvKfsEnahw4dUkpKitVhAABMys/PV9OmTQNy7NLSUrVoVkcFhS7Tx2rUqJHy8vJCKnGHTdKOi4uTJL3/SUPVqUOvvt1Nad/b6hAA+FmFyrVZb3r+Pw+EsrIyFRS6lJfbTM646ueKohNuteh+QGVlZSTt6jjbJV6nToTqmLgQCA+RjiirQwDgbz+/6SIYQ5zOuAhTSTtUhU3SBgCgqlyGWy4Tr8NyGW7/BeNHJG0AgO24Zcit6mdtM/sGkv36DgAAsCkqbQCA7bjllpkObnN7Bw5JGwBgOy7DkMuofhe3mX0Die5xAADCBJU2AMB27DoRjaQNALAdtwy5bJi06R4HACBMUGkDAGyH7nEAAMKEXWePk7QBALbj/nkxs38oYkwbAIAwQaUNALAdl8nZ42b2DSQqbQCA7bgM84svNm3apMGDBys5OVkOh0Nr1qyp9Jndu3frhhtuUHx8vOLi4tS7d28dPHjQp/OQtAEAMKm4uFidO3fWggULzrn966+/Vt++fdWmTRtt2LBBn3/+uaZPn66YmBifzkP3OADAdoI9ES09PV3p6enn3T5t2jRdd911euaZZzxtLVu29DkuKm0AgO245ZDLxOKWw3+xuN164403dMkll2jgwIFKTExUr169ztmF/ntI2gAAnEdRUZHXcvr0aZ+PUVhYqJMnT+rpp5/Wtddeq3feeUc33nijbrrpJm3cuNGnY5G0AQC24zbML5KUkpKi+Ph4z5KZmel7LO4zne1DhgzRgw8+qC5duuiRRx7RoEGDtHjxYp+OxZg2AMB2znZzm9lfkvLz8+V0Oj3t0dHRPh+rQYMGioyMVLt27bza27Ztq82bN/t0LJI2AADn4XQ6vZJ2ddSsWVM9e/bUnj17vNr37t2rZs2a+XQskjYAwHb8VWlX1cmTJ7Vv3z7Pel5enrZv366EhASlpqZqypQpGjZsmPr166f+/ftr/fr1eu2117RhwwafzkPSBgDYjttwyG1UP2n7um9OTo769+/vWc/IyJAkjRo1SsuWLdONN96oxYsXKzMzU5MmTVLr1q31yiuvqG/fvj6dh6QNALCdYFfaaWlpMn7nzWB333237r777mrHJDF7HACAsEGlDQCwHZci5DJRl7r8GIs/kbQBALZjmBzTNkzsG0h0jwMAECaotAEAthPsiWjBQtIGANiOy4iQyzAxpu3j+7SDhe5xAADCBJU2AMB23HLIbaIudSs0S22SNgDAduw6pk33OAAAYYJKGwBgO+YnotE9DgBAUJwZ0zbxwpAQ7R4naQMAbMdt8jGmoToRjTFtAADCBJU2AMB2GNMGACBMuBVhy/u06R4HACBMUGkDAGzHZTjkMvF6TTP7BhJJGwBgOy6Ts8dddI8DAAAzqLQBALbjNiLkNjF73M3scQAAgoPucQAAYCkqbQCA7bhlbga423+h+BVJGwBgO+YfrhKaHdEkbQCA7Zh/jGloJu3QjAoAAFRCpQ0AsB3ep42A2veJU+8+30QHv6ijosKaGvPCbnUeeNTrMwVfxWrt08217xOnDLdDjS8p0V1/+acSmpRZFDX8adCoI7rl/sNKSCzXgb0xWvxYsr78tI7VYSEAuNaBR/d4gCxcuFAtWrRQTEyMunfvrg8++MDqkCxxuiRCTdoW65Ynvj7n9sMHYvTszR2VdFGJJv3jSz2yfpsGTsxXVHRo3ksI31x5wzGNffyQVs5L1LgBl+jLT2rrqZfy1JBfyGyHaw0zLE3aq1at0uTJkzVt2jRt27ZNV1xxhdLT03Xw4EErw7JE+/4/adCUg+qSfvSc21//z1S1739MQ//jgFI6FKtB6ml1uOqY4hqUBzlSBMJN9x7R2ysTtH5FfeXvi9HiGU10+FCUBt3xo9Whwc+41sFx9uEqZpZQZGlUc+fO1ejRozVmzBi1bdtWWVlZSklJ0aJFi6wMK+S43dLO9xKU2OKU/nJ7O/2pW0/9eUgnff52gtWhwQ8io9xq1alEuRvjvNpzN8apXY9ii6JCIHCtg8dtOEwvociypF1WVqbc3FwNGDDAq33AgAH66KOPLIoqNJ08EqXTxTWUvaip2l75k8a/uEudBv6o/76vjb7a4rQ6PJjkTHCpRqT00xHvKSY/HY5UvcQKi6JCIHCtYZZlSfvIkSNyuVxKSkryak9KSlJBQYFFUYUm4+ff+Dpec1R/GHNITdsXa8C479T+qmPa/FIji6ODv/z6/QQOhxSijz+GSVzrwHOb7Br39eEqmzZt0uDBg5WcnCyHw6E1a9ac97P33XefHA6HsrKyfP5elnfaOxzeXRCGYVRqu9DVrleuiEi3GrUq8WpvdHGJjn0XbVFU8JeiozXkqpDqNfSutOIbVOjYYW7wsBOudfCcfcuXmcUXxcXF6ty5sxYsWPCbn1uzZo0++eQTJScnV+t7WZa0GzRooBo1alSqqgsLCytV3xe6yJqGmnU6qcJvYr3aC/NildDktEVRwV8qyiP01Y5a6tbvhFd7t34ntCuntkVRIRC41vaVnp6up556SjfddNN5P/Pdd99pwoQJeumllxQVFVWt81iWtGvWrKnu3bsrOzvbqz07O1uXXXaZRVFZ53RxhL7dWVvf7jzzD/fH/Bh9u7O2jn5XU5J01X3faevrDfThyiQd3h+jjcsa6cv/S9AVd3xvZdjwk1dfaKBrRxzVgOE/KuXiUt038zslNinXG3+rb3Vo8DOudXC45DC9SFJRUZHXcvp09Qolt9ut22+/XVOmTFH79u2r/b0s7Y/JyMjQ7bffrh49eqhPnz564YUXdPDgQY0dO9bKsCxxcEcdzRve0bO++skWkqRLb/5Bt//XPnW+9qiGzfpa2Qub6pUZLZR40SmNXvxPXdTzxPkOiTCycV09xdVzaeSDPyghsUIH9sTo0dtaqPDnX9pgH1zr4KhOF/ev95eklJQUr/YZM2Zo5syZPh9vzpw5ioyM1KRJk6odk2Rx0h42bJh+/PFHPfHEE/r+++/VoUMHvfnmm2rWrJmVYVmiVZ8izT/w4W9+ps+wQvUZVhikiBBsry9voNeXN7A6DAQB1zrwXJKnWq7u/pKUn58vp/Nfd+lER/s+jyg3N1fPPfectm7danrOluUzH8aNG6dx48ZZHQYAAJU4nU6vpF0dH3zwgQoLC5Wamuppc7lceuihh5SVlaX9+/dX+ViWJ20AAPzNX93j/nD77bfr6quv9mobOHCgbr/9dt11110+HYukDQCwnWC/MOTkyZPat2+fZz0vL0/bt29XQkKCUlNTVb++90TDqKgoNWrUSK1bt/bpPCRtAABMysnJUf/+/T3rGRkZkqRRo0Zp2bJlfjsPSRsAYDuGyfdpGz7um5aWJuPXj7r7Db6MY/8SSRsAYDu8TxsAAFiKShsAYDtmX68Zqq/mJGkDAGzn7Nu6zOwfikIzKgAAUAmVNgDAdugeBwAgTLgVIbeJzmQz+wYSSRsAYDsuwyGXiWrZzL6BFJq/SgAAgEqotAEAtsOYNgAAYcIw+ZYvgyeiAQAAM6i0AQC245JDLhMvDDGzbyCRtAEAtuM2zI1Lu6v+wq6gonscAIAwQaUNALAdt8mJaGb2DSSSNgDAdtxyyG1iXNrMvoFE0gYA2A5PRAMAAJai0gYA2A5j2gAAhAm3TD7GNETHtEPzVwkAAFAJlTYAwHYMk7PHjRCttEnaAADbsetbvugeBwAgTFBpAwBsh9njAACECbrHAQCApai0AQC2w7PHAQAIE3btHidpAwBsx65JmzFtAADCBEkbAGA7ZyttM4svNm3apMGDBys5OVkOh0Nr1qzxbCsvL9fUqVPVsWNH1a5dW8nJybrjjjt06NAhn78XSRsAYDvBTtrFxcXq3LmzFixYUGlbSUmJtm7dqunTp2vr1q169dVXtXfvXt1www0+fy/GtAEAMCk9PV3p6enn3BYfH6/s7Gyvtvnz5+vSSy/VwYMHlZqaWuXzkLQBALZjyNxtW4b/Qjmn48ePy+FwqG7duj7tR9IGANiOv2aPFxUVebVHR0crOjraVGylpaV65JFHNGLECDmdTp/2ZUwbAIDzSElJUXx8vGfJzMw0dbzy8nINHz5cbrdbCxcu9Hl/Km0AgO34q9LOz8/3qobNVNnl5eW69dZblZeXp/fee8/nKlsiaQMAbMhfSdvpdFYruf7a2YT91Vdf6f3331f9+vWrdRySNgAAJp08eVL79u3zrOfl5Wn79u1KSEhQcnKybr75Zm3dulWvv/66XC6XCgoKJEkJCQmqWbNmlc9D0gYA2E6wH2Oak5Oj/v37e9YzMjIkSaNGjdLMmTO1bt06SVKXLl289nv//feVlpZW5fOQtAEAtmMYDhkmkrav+6alpckwzn+j2G9t8wVJGwBgO3Z9NSe3fAEAECaotAEAtmPXV3OStAEAthPsMe1goXscAIAwQaUNALAduscBAAgTdI8DAABLhV2lPaV9b0U6oqwOAwH29qHtVoeAILqu8zVWh4AgMNxl0pEgnctk93ioVtphl7QBAPg9hiQzDyHzz/PL/I/ucQAAwgSVNgDAdtxyyGHDx5iStAEAtmPX2eMkbQCA7bgNhxw2vE+bMW0AAMIElTYAwHYMw+Ts8RCdPk7SBgDYjl3HtOkeBwAgTFBpAwBsx66VNkkbAGA7zB4HAACWotIGANgOs8cBAAgTZ5K2mTFtPwbjR3SPAwAQJqi0AQC2w+xxAADChCFz78QO0d5xkjYAwH7sWmkzpg0AQJig0gYA2I9N+8dJ2gAA+zHZPS66xwEAgBlU2gAA2+GJaAAAhAlmjwMAgHPatGmTBg8erOTkZDkcDq1Zs8Zru2EYmjlzppKTkxUbG6u0tDTt3LnT5/OQtAEA9mM4zC8+KC4uVufOnbVgwYJzbn/mmWc0d+5cLViwQJ999pkaNWqka665RidOnPDpPHSPAwBsJ9hj2unp6UpPTz/PsQxlZWVp2rRpuummmyRJy5cvV1JSklasWKH77ruvyueh0gYA4DyKioq8ltOnT/t8jLy8PBUUFGjAgAGetujoaF155ZX66KOPfDoWSRsAYD+GHxZJKSkpio+P9yyZmZk+h1JQUCBJSkpK8mpPSkrybKuqKnWPz5s3r8oHnDRpkk8BAADgb/6aPZ6fny+n0+lpj46OrvYxHQ7veAzDqNT2e6qUtJ999tkqB0TSBgCEBD/ca+10Or2SdnU0atRI0pmKu3Hjxp72wsLCStX376lS0s7Ly/PpoAAA4IwWLVqoUaNGys7OVteuXSVJZWVl2rhxo+bMmePTsao9e7ysrEx5eXm66KKLFBnJJHQAQOgI9sNVTp48qX379nnW8/LytH37diUkJCg1NVWTJ0/W7Nmz1apVK7Vq1UqzZ89WrVq1NGLECJ/O43O2LSkp0cSJE7V8+XJJ0t69e9WyZUtNmjRJycnJeuSRR3w9JAAA/hXkt3zl5OSof//+nvWMjAxJ0qhRo7Rs2TI9/PDDOnXqlMaNG6djx46pV69eeueddxQXF+fTeXyePf6nP/1Jn3/+uTZs2KCYmBhP+9VXX61Vq1b5ejgAAMJeWlqaDMOotCxbtkzSmTlfM2fO1Pfff6/S0lJt3LhRHTp08Pk8Plfaa9as0apVq9S7d2+vWW/t2rXT119/7XMAAAD4n+Pnxcz+ocfnpH348GElJiZWai8uLvZ56joAAAER5O7xYPG5e7xnz5564403POtnE/WSJUvUp08f/0UGAAC8+FxpZ2Zm6tprr9WuXbtUUVGh5557Tjt37tTHH3+sjRs3BiJGAAB8Q6V9xmWXXaYPP/xQJSUluuiii/TOO+8oKSlJH3/8sbp37x6IGAEA8E2Q3/IVLNW6wbpjx46eW74AAEBwVCtpu1wurV69Wrt375bD4VDbtm01ZMgQHrICAAgJwX41Z7D4nGW//PJLDRkyRAUFBWrdurWkMw9YadiwodatW6eOHTv6PUgAAHzCmPYZY8aMUfv27fXtt99q69at2rp1q/Lz89WpUyfde++9gYgRAADfMKZ9xueff66cnBzVq1fP01avXj3NmjVLPXv29GtwAADgX3yutFu3bq0ffvihUnthYaEuvvhivwQFAIAZDsP8EoqqVGkXFRV5/jx79mxNmjRJM2fOVO/evSVJW7Zs0RNPPOHzK8YAAAgIm45pVylp161b1+sRpYZh6NZbb/W0GT9Psxs8eLBcLlcAwgQAAFVK2u+//36g4wAAwH/MTiYL54loV155ZaDjAADAfy7k7vFzKSkp0cGDB1VWVubV3qlTJ9NBAQCAyqr1as677rpLb7311jm3M6YNALCcTSttn2/5mjx5so4dO6YtW7YoNjZW69ev1/Lly9WqVSutW7cuEDECAOAbww9LCPK50n7vvfe0du1a9ezZUxEREWrWrJmuueYaOZ1OZWZm6vrrrw9EnAAAXPB8rrSLi4uVmJgoSUpISNDhw4clnXnz19atW/0bHQAA1cFjTM9o3bq19uzZo+bNm6tLly56/vnn1bx5cy1evFiNGzcORIwXrEGjjuiW+w8rIbFcB/bGaPFjyfry0zpWhwWTvthSW/9vYaK++qKWjv4QpRn/nafL0o97tg9M7nLO/cY8+p1uGXc4SFEiEDp0O6Z/u/OALm5bpPqJZXpycid9/H6i1WHZktmnmoXqE9GqNab9/fffS5JmzJih9evXKzU1VfPmzdPs2bN9OtamTZs0ePBgJScny+FwaM2aNb6GY1tX3nBMYx8/pJXzEjVuwCX68pPaeuqlPDVsUvb7OyOklZZEqGX7Uxo/69tzbl+5/UuvJWPuQTkchvpef/ycn0f4iIl1KW9PHS16uo3VodgfY9pnjBw50vPnrl27av/+/frnP/+p1NRUNWjQwKdjFRcXq3Pnzrrrrrv0b//2b76GYms33XtEb69M0PoV9SVJi2c0Ufe0Exp0x49amkmPRjjr+YcT6vmHE+fdnpBY4bX+8dvx6nz5STVuxi9s4S7nwwbK+dC3/yeBX6r2fdpn1apVS926davWvunp6UpPTzcbgu1ERrnVqlOJVi3w7jbL3Rindj2KLYoKVjh2OFKfvuvUH7MOWB0KgBBQpaSdkZFR5QPOnTu32sHgDGeCSzUipZ+OeF+enw5Hqt6vqjDYW/bLCYqt41Lf6+gaB3zhkMkxbb9F4l9VStrbtm2r0sF++VIRmGf86i+cw6GQHWdBYLz9jwT94cZjqhnDhQfAC0NCUtHRGnJVSPUaelfV8Q0qdOyw6RENhIkvPqmtb7+O0X8s3m91KED4sekLQ3yePY7AqyiP0Fc7aqlbP+/JSt36ndCunNoWRYVge3tlfbXqVKKL2pdaHQoQfpg9jmB69YUGmjIvX3t3xGp3Tm1dd9uPSmxSrjf+Vt/q0GDSqeIIHcqL9qwX5NfU11/GKq5uhRKblkuSik9EaNNr8bp3xiGrwkQAxMRWKDn1lGc9qckptWx9QieOR+lwQYyFkSFcWJq0T548qX379nnW8/LytH37diUkJCg1NdXCyKy3cV09xdVzaeSDPyghsUIH9sTo0dtaqPC7mlaHBpP2fl5LD998sWf9+ZlNJEnX3HpUf8w6KEnauLaeZDjUf+gxS2JEYLRqX6Q5//2vJ0feO+UrSVL22sZ69rH2VoVlTzZ9YYjDMH493Sl4NmzYoP79+1dqHzVqlJYtW+bVVlRUpPj4eKVpiCIdUUGKEFZ5+9B2q0NAEF3X+RqrQ0AQVLjL9O6R/9bx48fldDoDco6zuaL5rFmKiKl+74W7tFT7p00LaKzVYWmlnZaWJgt/ZwAAIKxUayLaiy++qMsvv1zJyck6cODMQx+ysrK0du1avwYHAEC1BHkiWkVFhR599FG1aNFCsbGxatmypZ544gm53W7/fJ+f+Zy0Fy1apIyMDF133XX66aef5HK5JEl169ZVVlaWX4MDAKBagpy058yZo8WLF2vBggXavXu3nnnmGf3nf/6n5s+f75/v8zOfk/b8+fO1ZMkSTZs2TTVq1PC09+jRQ1988YVfgwMAoDrOvuXLzOKLjz/+WEOGDNH111+v5s2b6+abb9aAAQOUk5Pj1+/lc9LOy8tT165dK7VHR0eruJjnYgMALjx9+/bVu+++q71790qSPv/8c23evFnXXXedX8/j80S0Fi1aaPv27WrWrJlX+1tvvaV27dr5LTAAAKrNT09EKyoq8mqOjo5WdHR0pY9PnTpVx48fV5s2bVSjRg25XC7NmjVL//7v/179GM7B56Q9ZcoUjR8/XqWlpTIMQ59++qlWrlypzMxM/fWvf/VrcAAAVIuf7tNOSUnxap4xY4ZmzpxZ6eOrVq3S3//+d61YsULt27fX9u3bNXnyZCUnJ2vUqFEmAvHmc9K+6667VFFRoYcfflglJSUaMWKEmjRpoueee07Dhw/3W2AAAFgtPz/f6z7tc1XZ0pmC9pFHHvHkwY4dO+rAgQPKzMy0NmlL0j333KN77rlHR44ckdvtVmJi4u/vBABAkFRnMtmv95ckp9NZpYerlJSUKCLCe5pYjRo1/H7Ll6mHqzRo0MBfcQAA4D9Bfozp4MGDNWvWLKWmpqp9+/batm2b5s6dq7vvvttEEJVVayLab703+5tvvjEVEAAA4Wb+/PmaPn26xo0bp8LCQiUnJ+u+++7TY4895tfz+Jy0J0+e7LVeXl6ubdu2af369ZoyZYq/4gIAoPpMdo/7WmnHxcUpKysr4A8Z8zlpP/DAA+ds/8tf/uL3m8gBAKgWm77lq1rPHj+X9PR0vfLKK/46HAAA+BW/veXrf//3f5WQkOCvwwEAUH02rbR9Ttpdu3b1mohmGIYKCgp0+PBhLVy40K/BAQBQHf665SvU+Jy0hw4d6rUeERGhhg0bKi0tTW3atPFXXAAA4Fd8StoVFRVq3ry5Bg4cqEaNGgUqJgAAcA4+TUSLjIzU/fffr9OnTwcqHgAAzAvy+7SDxefZ47169dK2bdsCEQsAAH4R7PdpB4vPY9rjxo3TQw89pG+//Vbdu3dX7dq1vbZ36tTJb8EBAIB/qXLSvvvuu5WVlaVhw4ZJkiZNmuTZ5nA4ZBiGHA6HXC6X/6MEAMBXIVotm1HlpL18+XI9/fTTysvLC2Q8AACYd6Hfp20YZ75Bs2bNAhYMAAA4P5/GtH/r7V4AAIQKHq4i6ZJLLvndxH306FFTAQEAYNqF3j0uSY8//rji4+MDFQsAAPgNPiXt4cOHKzExMVCxAADgFxd89zjj2QCAsHGhd4+fnT0OAEDIu9CTttvtDmQcAADgd/j8GFMAAELdBT+mDQBA2LBp97jPb/kCAADWoNIGANiPTSttkjYAwHbsOqZN9zgAAGGCShsAYD90jwMAEB7oHgcAAJai0gYA2A/d4wAAhAmSNgAA4cHx82Jm/1DEmDYAAGGCShsAYD827R6n0gYA2M7ZW77MLL767rvvdNttt6l+/fqqVauWunTpotzcXL9+LyptAABMOnbsmC6//HL1799fb731lhITE/X111+rbt26fj0PSRsAYD9B7h6fM2eOUlJStHTpUk9b8+bNTQRwbnSPAwDsyTCx/KyoqMhrOX369DlPtW7dOvXo0UO33HKLEhMT1bVrVy1ZssTvX4mkDQDAeaSkpCg+Pt6zZGZmnvNz33zzjRYtWqRWrVrp7bff1tixYzVp0iT97W9/82s8dI8DAGzHX88ez8/Pl9Pp9LRHR0ef8/Nut1s9evTQ7NmzJUldu3bVzp07tWjRIt1xxx3VD+RXqLQBAPZjpmv8F13kTqfTazlf0m7cuLHatWvn1da2bVsdPHjQr1+LpA0AgEmXX3659uzZ49W2d+9eNWvWzK/nIWkDAGwn2PdpP/jgg9qyZYtmz56tffv2acWKFXrhhRc0fvx4v34vkjYAwH781D1eVT179tTq1au1cuVKdejQQU8++aSysrI0cuRI/3yfnzERDQBgO/6aiOaLQYMGadCgQdU/aRWQtBGSBiZ3sToEBFHh2vpWh4AgcJWcloZbHUV4I2kDAOzHpi8MIWkDAOzHpkmbiWgAAIQJKm0AgO1YMREtGEjaAAD7oXscAABYiUobAGA7DsOQw6h+uWxm30AiaQMA7IfucQAAYCUqbQCA7TB7HACAcGHT7nGSNgDAduxaaTOmDQBAmKDSBgDYD93jAACEB7rHAQCApai0AQD2Q/c4AADhI1S7uM2gexwAgDBBpQ0AsB/DOLOY2T8EkbQBALbD7HEAAGApKm0AgP0wexwAgPDgcJ9ZzOwfikjaAAD7sWmlzZg2AABhgkobAGA7dp09TtIGANiPTe/TpnscAIAwQaUNALAduscBAAgXzB4HAABVkZmZKYfDocmTJ/v1uFTaAADbsbJ7/LPPPtMLL7ygTp06Vf8g50GlDQCwn7Ozx80s1XDy5EmNHDlSS5YsUb169fz8pUjaAAD4zfjx43X99dfr6quvDsjx6R4HANiOv7rHi4qKvNqjo6MVHR19zn3+8Y9/aOvWrfrss8+qf+LfQaUNALAfww+LpJSUFMXHx3uWzMzMc54uPz9fDzzwgP7+978rJiYmYF+LShsAYDv+qrTz8/PldDo97eersnNzc1VYWKju3bt72lwulzZt2qQFCxbo9OnTqlGjRvUD+hlJGwCA83A6nV5J+3yuuuoqffHFF15td911l9q0aaOpU6f6JWFLJG0AgB25jTOLmf19EBcXpw4dOni11a5dW/Xr16/UbgZJGwBgPzZ9IhpJGwCAANiwYYPfj0nSBgDYjkMmJ6L5LRL/ImkDAOyH92kDAAArUWkDAGyH92kDABAubDp7nO5xAADCBJU2AMB2HIYhh4nJZGb2DSSSNgDAftw/L2b2D0EkbQCA7di10mZMGwCAMEGlDQCwH5vOHidpAwDshyeiAQAAK1Fph7BBo47olvsPKyGxXAf2xmjxY8n68tM6VoeFAOF620/UzhLVWv2jIvedVo1jFfrpT01U1jvOsz364xOKWf+Tor4uVcQJl44+21wVLWMsjNg+7PpENCrtEHXlDcc09vFDWjkvUeMGXKIvP6mtp17KU8MmZVaHhgDgetuTo9StiuYxOnlf0nm3l7eN1ck7GgY5sgvA2e5xM0sIsjRpZ2ZmqmfPnoqLi1NiYqKGDh2qPXv2WBlSyLjp3iN6e2WC1q+or/x9MVo8o4kOH4rSoDt+tDo0BADX257KutdR8W0NdbpP3Dm3l/aPV8nwBirrXCvIkSFcWZq0N27cqPHjx2vLli3Kzs5WRUWFBgwYoOLiYivDslxklFutOpUod6P3P/TcjXFq1+PC/tnYEdcb8D+H2/wSiiwd016/fr3X+tKlS5WYmKjc3Fz169fPoqis50xwqUak9NMR78vz0+FI1UussCgqBArXGwgAZo8H3vHjxyVJCQkJFkcSGn79d8bhUMjeOwjzuN4Afk/IzB43DEMZGRnq27evOnToYHU4lio6WkOuCqleQ+8qK75BhY4dDplLBj/hegMBYNOHq4RMpT1hwgTt2LFDK1eutDoUy1WUR+irHbXUrd8Jr/Zu/U5oV05ti6JCoHC9Af87++xxM0soColf4ydOnKh169Zp06ZNatq0qdXhhIRXX2igKfPytXdHrHbn1NZ1t/2oxCbleuNv9a0ODQHA9bYnxym3anz/r9v2avxQrshvSuWOqyF3wyg5TrhU43C5Io6e6WWp8d2Zz7rrRcpdLyT+ew5fNh3TtvRvhWEYmjhxolavXq0NGzaoRYsWVoYTUjauq6e4ei6NfPAHJSRW6MCeGD16WwsVflfT6tAQAFxve4rcd0r1Hs33rMf9T6Ek6dQfnDrxQLKiPz0h57wCz/b4Px+SJBUPr6/if+febVRmadIeP368VqxYobVr1youLk4FBWf+8sbHxys2NtbK0ELC68sb6PXlDawOA0HC9baf8o61Vbi2zXm3l15VV6VX1Q1eQBcSQ+beiR2ahba1Y9qLFi3S8ePHlZaWpsaNG3uWVatWWRkWACDMMaYdAEaI/lAAAAhFzHQAANiPIZMT0fwWiV+RtAEA9mPT2eMhc582AAD4bVTaAAD7cUtymNw/BJG0AQC2Y3YGOLPHAQAIFsa0AQCAlUjaAAD7OVtpm1l8kJmZqZ49eyouLk6JiYkaOnSo9uzZ4/evRdIGANhPkJP2xo0bNX78eG3ZskXZ2dmqqKjQgAEDVFxc7NevxZg2AAAmrV+/3mt96dKlSkxMVG5urvr16+e385C0AQD246dbvoqKiryao6OjFR0d/bu7Hz9+XJKUkJBgIojK6B4HANiOv14YkpKSovj4eM+SmZn5u+c2DEMZGRnq27evOnTo4NfvRaUNAMB55Ofny+l0etarUmVPmDBBO3bs0ObNm/0eD0kbAGA/frpP2+l0eiXt3zNx4kStW7dOmzZtUtOmTat//vMgaQMA7MdtSA4TSdvt276GYWjixIlavXq1NmzYoBYtWlT/3L+BpA0AgEnjx4/XihUrtHbtWsXFxamgoECSFB8fr9jYWL+dh4loAAD7CfJ92osWLdLx48eVlpamxo0be5ZVq1b59WtRaQMAbMjkmLZ87x4PBpI2AMB+eGEIAACwEpU2AMB+3IZ87eKuvH/oIWkDAOzHcJ9ZzOwfgugeBwAgTFBpAwDsx6YT0UjaAAD7semYNt3jAACECSptAID90D0OAECYMGQyafstEr+iexwAgDBBpQ0AsB+6xwEACBNutyQTD0hxh+bDVUjaAAD7sWmlzZg2AABhgkobAGA/Nq20SdoAAPvhiWgAAMBKVNoAANsxDLcME6/XNLNvIJG0AQD2YxjmurhDdEyb7nEAAMIElTYAwH4MkxPRQrTSJmkDAOzH7ZYcJsalQ3RMm+5xAADCBJU2AMB+6B4HACA8GG63DBPd49zyBQBAsNi00mZMGwCAMEGlDQCwH7chOexXaZO0AQD2YxiSzNzyFZpJm+5xAADCBJU2AMB2DLchw0T3uEGlDQBAkBhu80s1LFy4UC1atFBMTIy6d++uDz74wK9fi6QNAIAfrFq1SpMnT9a0adO0bds2XXHFFUpPT9fBgwf9dg6SNgDAdgy3YXrx1dy5czV69GiNGTNGbdu2VVZWllJSUrRo0SK/fS+SNgDAfoLcPV5WVqbc3FwNGDDAq33AgAH66KOP/Pa1wmYi2tlJARUqN/WQGwChx1Vy2uoQEARnr3MwJnmZzRUVKpckFRUVebVHR0crOjq60uePHDkil8ulpKQkr/akpCQVFBRUP5BfCZukfeLECUnSZr1pcSQA/G641QEgmE6cOKH4+PiAHLtmzZpq1KiRNheYzxV16tRRSkqKV9uMGTM0c+bM8+7jcDi81g3DqNRmRtgk7eTkZOXn5ysuLs6vPwAAQHAYhqETJ04oOTk5YOeIiYlRXl6eysrKTB/rXAn3XFW2JDVo0EA1atSoVFUXFhZWqr7NCJukHRERoaZNm1odBgDAhEBV2L8UExOjmJiYgJ/nl2rWrKnu3bsrOztbN954o6c9OztbQ4YM8dt5wiZpAwAQyjIyMnT77berR48e6tOnj1544QUdPHhQY8eO9ds5SNoAAPjBsGHD9OOPP+qJJ57Q999/rw4dOujNN99Us2bN/HYObvkKYYF+sg5Cw6ZNmzR48GAlJyfL4XBozZo1VoeEAMnMzFTPnj0VFxenxMREDR06VHv27LE6LPjRuHHjtH//fp0+fVq5ubnq16+fX49P0g5RwXiyDkJDcXGxOnfurAULFlgdCgJs48aNGj9+vLZs2aLs7GxVVFRowIABKi4utjo0hAmHEapPRb/A9erVS926dfN6kk7btm01dOhQZWZmWhgZAsnhcGj16tUaOnSo1aEgCA4fPqzExERt3LjR7xUZ7IlKOwQF68k6AKx1/PhxSVJCQoLFkSBckLRDULCerAPAOoZhKCMjQ3379lWHDh2sDgdhgtnjISzQT9YBYJ0JEyZox44d2rx5s9WhIIyQtENQsJ6sA8AaEydO1Lp167Rp0yYeGgWf0D0egn75ZJ1fys7O1mWXXWZRVADMMgxDEyZM0Kuvvqr33ntPLVq0sDokhBkq7RAVjCfrIDScPHlS+/bt86zn5eVp+/btSkhIUGpqqoWRwd/Gjx+vFStWaO3atYqLi/P0psXHxys2Ntbi6BAOuOUrhC1cuFDPPPOM58k6zz77LLeF2NCGDRvUv3//Su2jRo3SsmXLgh8QAuZ8c1KWLl2qO++8M7jBICyRtAEACBOMaQMAECZI2gAAhAmSNgAAYYKkDQBAmCBpAwAQJkjaAACECZI2AABhgqQNAECYIGkDPpg5c6a6dOniWb/zzjs1dOjQoMexf/9+ORwObd++/byfad68ubKysqp8zGXLlqlu3bqmY3M4HFqzZo3p4wCojKSNsHfnnXfK4XDI4XAoKipKLVu21B//+EcVFxcH/NzPPfdclR81WpVECwC/hReGwBauvfZaLV26VOXl5frggw80ZswYFRcXa9GiRZU+W15erqioKL+cNz4+3i/HAYCqoNKGLURHR6tRo0ZKSUnRiBEjNHLkSE8X7dku7f/5n/9Ry5YtFR0dLcMwdPz4cd17771KTEyU0+nUH/7wB33++edex3366aeVlJSkuLg4jR49WqWlpV7bf9097na7NWfOHF188cWKjo5WamqqZs2aJUme1zB27dpVDodDaWlpnv2WLl2qtm3bKiYmRm3atNHChQu9zvPpp5+qa9euiomJUY8ePbRt2zaff0Zz585Vx44dVbt2baWkpGjcuHE6efJkpc+tWbNGl1xyiWJiYnTNNdcoPz/fa/trr72m7t27KyYmRi1bttTjjz+uiooKn+MB4DuSNmwpNjZW5eXlnvV9+/bp5Zdf1iuvvOLpnr7++utVUFCgN998U7m5uerWrZuuuuoqHT16VJL08ssva8aMGZo1a5ZycnLUuHHjSsn01/70pz9pzpw5mj59unbt2qUVK1YoKSlJ0pnEK0n/93//p++//16vvvqqJGnJkiWaNm2aZs2apd27d2v27NmaPn26li9fLkkqLi7WoEGD1Lp1a+Xm5mrmzJn64x//6PPPJCIiQvPmzdOXX36p5cuX67333tPDDz/s9ZmSkhLNmjVLy5cv14cffqiioiINHz7cs/3tt9/WbbfdpkmTJmnXrl16/vnntWzZMs8vJgACzADC3KhRo4whQ4Z41j/55BOjfv36xq233moYhmHMmDHDiIqKMgoLCz2feffddw2n02mUlpZ6Heuiiy4ynn/+ecMwDKNPnz7G2LFjvbb36tXL6Ny58znPXVRUZERHRxtLliw5Z5x5eXmGJGPbtm1e7SkpKcaKFSu82p588kmjT58+hmEYxvPPP28kJCQYxcXFnu2LFi0657F+qVmzZsazzz573u0vv/yyUb9+fc/60qVLDUnGli1bPG27d+82JBmffPKJYRiGccUVVxizZ8/2Os6LL75oNG7c2LMuyVi9evV5zwug+hjThi28/vrrqlOnjioqKlReXq4hQ4Zo/vz5nu3NmjVTw4YNPeu5ubk6efKk6tev73WcU6dO6euvv5Yk7d69W2PHjvXa3qdPH73//vvnjGH37t06ffq0rrrqqirHffjwYeXn52v06NG65557PO0VFRWe8fLdu3erc+fOqlWrllccvnr//fc1e/Zs7dq1S0VFRaqoqFBpaamKi4tVu3ZtSVJkZKR69Ojh2adNmzaqW7eudu/erUsvvVS5ubn67LPPvCprl8ul0tJSlZSUeMUIwP9I2rCF/v37a9GiRYqKilJycnKliWZnk9JZbrdbjRs31oYNGyodq7q3PcXGxvq8j9vtlnSmi7xXr15e22rUqCFJMvzwyvsDBw7ouuuu09ixY/Xkk08qISFBmzdv1ujRo72GEaQzt2z92tk2t9utxx9/XDfddFOlz8TExJiOE8BvI2nDFmrXrq2LL764yp/v1q2bCgoKFBkZqebNm5/zM23bttWWLVt0xx13eNq2bNly3mO2atVKsbGxevfddzVmzJhK22vWrCnpTGV6VlJSkpo0aaJvvvlGI0eOPOdx27VrpxdffFGnTp3y/GLwW3GcS05OjioqKvRf//Vfiog4M5Xl5ZdfrvS5iooK5eTk6NJLL5Uk7dmzRz/99JPatGkj6czPbc+ePT79rAH4D0kbF6Srr75affr00dChQzVnzhy1bt1ahw4d0ptvvqmhQ4eqR48eeuCBBzRq1Cj16NFDffv21UsvvaSdO3eqZcuW5zxmTEyMpk6dqocfflg1a9bU5ZdfrsOHD2vnzp0aPXq0EhMTFRsbq/Xr16tp06aKiYlRfHy8Zs6cqUmTJsnpdCo9PV2nT59WTk6Ojh07poyMDI0YMULTpk3T6NGj9eijj2r//v3685//7NP3veiii1RRUaH58+dr8ODB+vDDD7V48eJKn4uKitLEiRM1b948RUVFacKECerdu7cniT/22GMaNGiQUlJSdMsttygiIkI7duzQF198oaeeesr3CwHAJ8wexwXJ4XDozTffVL9+/XT33Xfrkksu0fDhw7V//37PbO9hw4bpscce09SpU9W9e3cdOHBA999//28ed/r06XrooYf02GOPqW3btho2bJgKCwslnRkvnjdvnp5//nklJydryJAhkqQxY8bor3/9q5YtW6aOHTvqyiuv1LJlyzy3iNWpU0evvfaadu3apa5du2ratGmaM2eOT9+3S5cumjt3rubMmaMOHTropZdeUmZmZqXP1apVS1OnTtWIESPUp08fxcbG6h//+Idn+8CBA/X6668rOztbPXv2VO/evTV37lw1a9bMp3gAVI/D8MeAGQAACDgqbQAAwgRJGwCAMEHSBgAgTJC0AQAIEyRtAADCBEkbAIAwQdIGACBMkLQBAAgTJG0AAMIESRsAgDBB0gYAIEyQtAEACBP/H5gmosubXJN4AAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "disp = ConfusionMatrixDisplay(confusao)\n",
    "disp.plot()\n",
    "plt.tick_params(axis=u'both', which=u'both',length=0)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "f49c5d1e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_model(modelo, show_shapes=True, show_dtype=True, show_layer_names=False, expand_nested=True, show_layer_activations=True)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}