Deep_Learning.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/salimNabbout/MFP/blob/main/Deep_Learning.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "sNc3NouWYrN7"
      },
      "cell_type": "markdown",
      "source": [
        "Imports"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##POS GRADUAÇÃO EM CIÊNCIAS DE DADOS - PUC RIO\n",
        "\n",
        "##MVP SPRINT 2\n",
        "\n",
        "##PROJETO - Deep Learning\n",
        "\n",
        "##ALUNO: SALIM JORGE NABBOUT\n",
        "\n"
      ],
      "metadata": {
        "id": "Z9n-KE8wfuXx"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 1.0 Justificativa do Projeto\n",
        "\n",
        "\n",
        "O projeto consiste modelo baseado em Deep Learning para a identificação de imagens oriundas de uma determinado Banco de dados.\n",
        "\n",
        "para este projeto foi utilizado o dataset CIFAR 10\n",
        "\n",
        "O conjunto de dados CIFAR-10 consiste em 60.000 imagens coloridas 32x32 em 10 classes, com 6.000 imagens por classe. Existem 50.000 imagens de treinamento e 10.000 imagens de teste.\n",
        "\n",
        "Premissas:\n",
        "\n",
        "As classes são completamente mutuamente exclusivas. Não há sobreposição entre automóveis e caminhões. \"Automóvel\" inclui sedans, SUVs, coisas desse tipo. \"Caminhão\" inclui apenas caminhões grandes.\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "XFpMGSYYsL2c"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 2.0 Importanto Bibliotecas"
      ],
      "metadata": {
        "id": "u3Yk_VwQf-nB"
      }
    },
    {
      "metadata": {
        "id": "4bQCahYjYdUB"
      },
      "cell_type": "code",
      "source": [
        "import tensorflow\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from tensorflow.keras.models import load_model\n",
        "import tensorflow as keras\n",
        "\n",
        "\n"
      ],
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 3.0 Carregando Dataset"
      ],
      "metadata": {
        "id": "1l-KhqOYgXBx"
      }
    },
    {
      "metadata": {
        "id": "QoIys3wMYirH"
      },
      "cell_type": "code",
      "source": [
        "dataset = tf.keras.datasets.cifar10\n",
        "((imagens_treino, identificacoes_treino), (imagens_teste, identificacoes_teste)) = dataset.load_data()\n"
      ],
      "execution_count": 47,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 3.1 Detalhando DATASET\n",
        "\n",
        "0 - Avião\n",
        "\n",
        "1 - Automóvel\n",
        "\n",
        "2 - Passaro\n",
        "\n",
        "3 - Gato\n",
        "\n",
        "4 - Cervo\n",
        "\n",
        "5 - Cachorro\n",
        "\n",
        "6 - Sapo\n",
        "\n",
        "7 - Cavalo\n",
        "\n",
        "8 - Enviar\n",
        "\n",
        "9 - Caminhão"
      ],
      "metadata": {
        "id": "ICU5ddufuuif"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "#4.0 Explorando os dados"
      ],
      "metadata": {
        "id": "s1TjcqqyghCh"
      }
    },
    {
      "metadata": {
        "id": "fPIT36hpYl5A",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f757ce39-b23d-46ed-8263-74138acd5113"
      },
      "cell_type": "code",
      "source": [
        "len(imagens_treino)\n",
        "imagens_treino.shape\n",
        "imagens_teste.shape\n",
        "len(identificacoes_teste)\n",
        "identificacoes_treino.min()\n",
        "identificacoes_treino.max()"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "9"
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 5.0 Explorando os dados\n",
        "\n",
        "\n",
        "Observou se inconsistencia entre os nomes das classificações, oriundas do dataset e a imagem processada pelo plt.colorbar.\n",
        "\n",
        "As imagens treinos declaradas não conferem com as do dataset"
      ],
      "metadata": {
        "id": "GpiNq9_IgprJ"
      }
    },
    {
      "metadata": {
        "id": "bGESm49JVahh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 447
        },
        "outputId": "2b179440-51ff-4664-868b-44a42e3a2d3f"
      },
      "cell_type": "code",
      "source": [
        "total_de_classificacoes = 10\n",
        "nomes_de_classificacoes = ['Avião', 'Automóvel', 'Passaro',\n",
        "                           'Gato', 'Cervo', 'Cachorro', 'Sapo',\n",
        "                           'Cavalo', 'Enviar', 'Caminhão']\n",
        "\n",
        "plt.imshow(imagens_treino[1])\n",
        "plt.colorbar()"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.colorbar.Colorbar at 0x79e333462620>"
            ]
          },
          "metadata": {},
          "execution_count": 72
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "id": "uU6ZySpIcoc5"
      },
      "cell_type": "markdown",
      "source": [
        "Normalizando as imagens"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 6.0 Normalizando Imagens"
      ],
      "metadata": {
        "id": "_1ycZZAcgvO4"
      }
    },
    {
      "metadata": {
        "id": "Ral_hdl9ulGG"
      },
      "cell_type": "code",
      "source": [
        "imagens_treino = imagens_treino/float(255)"
      ],
      "execution_count": 73,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 7.0 Criando - Compilando - Treinando e Normalizando o Modelo"
      ],
      "metadata": {
        "id": "GVt7JZ72g9-Q"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "\n",
        "# Carregando o dataset do CIFAR 10\n",
        "(train_images, train_labels), (test_images, test_labels) = keras.datasets.cifar10.load_data()\n",
        "\n",
        "# Normalizando os valores de pixel para o intervalo [0, 1]\n",
        "train_images, test_images = train_images / 255.0, test_images / 255.0\n",
        "\n",
        "# Criando os Modelo\n",
        "modelo = keras.Sequential([\n",
        "    keras.layers.Flatten(input_shape=(32, 32, 3)),\n",
        "    keras.layers.Dense(256, activation=tf.nn.relu),\n",
        "    keras.layers.Dropout(0.2),\n",
        "    keras.layers.Dense(10, activation=tf.nn.softmax)\n",
        "])\n",
        "\n",
        "adam = keras.optimizers.Adam(lr=0.005)\n",
        "\n",
        "parando_cedo = [\n",
        "    keras.callbacks.EarlyStopping(monitor='val_loss'),\n",
        "    keras.callbacks.ModelCheckpoint(filepath='melhor_modelo.hdf5',\n",
        "                                    monitor='val_loss', save_best_only=True)\n",
        "]\n",
        "\n",
        "modelo.compile(optimizer=adam,\n",
        "               loss='sparse_categorical_crossentropy',\n",
        "               metrics=['accuracy'])\n",
        "\n",
        "# Variando Parametros - Melhora da Performance\n",
        "# Batch_Size\n",
        "# Epochs\n",
        "\n",
        "historico = modelo.fit(train_images, train_labels,\n",
        "                       batch_size=380,\n",
        "                       epochs=10,\n",
        "                       validation_split=0.2,\n",
        "                       callbacks=parando_cedo)\n"
      ],
      "metadata": {
        "id": "WEUyHe6UnVSb",
        "outputId": "bb2c913f-aead-4fa6-fbe3-1bacaadfc51e",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:`lr` is deprecated in Keras optimizer, please use `learning_rate` or use the legacy optimizer, e.g.,tf.keras.optimizers.legacy.Adam.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "106/106 [==============================] - 4s 35ms/step - loss: 2.2123 - accuracy: 0.2259 - val_loss: 1.9258 - val_accuracy: 0.3188\n",
            "Epoch 2/10\n",
            "106/106 [==============================] - 4s 39ms/step - loss: 1.9100 - accuracy: 0.3089 - val_loss: 1.8278 - val_accuracy: 0.3520\n",
            "Epoch 3/10\n",
            "106/106 [==============================] - 3s 32ms/step - loss: 1.8469 - accuracy: 0.3370 - val_loss: 1.8109 - val_accuracy: 0.3521\n",
            "Epoch 4/10\n",
            "106/106 [==============================] - 3s 33ms/step - loss: 1.8115 - accuracy: 0.3503 - val_loss: 1.7689 - val_accuracy: 0.3707\n",
            "Epoch 5/10\n",
            "106/106 [==============================] - 4s 38ms/step - loss: 1.7859 - accuracy: 0.3609 - val_loss: 1.7610 - val_accuracy: 0.3730\n",
            "Epoch 6/10\n",
            "106/106 [==============================] - 3s 33ms/step - loss: 1.7709 - accuracy: 0.3636 - val_loss: 1.7464 - val_accuracy: 0.3788\n",
            "Epoch 7/10\n",
            "106/106 [==============================] - 3s 33ms/step - loss: 1.7563 - accuracy: 0.3697 - val_loss: 1.7083 - val_accuracy: 0.3929\n",
            "Epoch 8/10\n",
            "106/106 [==============================] - 4s 36ms/step - loss: 1.7314 - accuracy: 0.3802 - val_loss: 1.6918 - val_accuracy: 0.4026\n",
            "Epoch 9/10\n",
            "106/106 [==============================] - 4s 36ms/step - loss: 1.7229 - accuracy: 0.3823 - val_loss: 1.6856 - val_accuracy: 0.3996\n",
            "Epoch 10/10\n",
            "106/106 [==============================] - 3s 32ms/step - loss: 1.7151 - accuracy: 0.3864 - val_loss: 1.6708 - val_accuracy: 0.4123\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 8.0 Sumário do Modelo"
      ],
      "metadata": {
        "id": "insLKnWkhyKh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sumario_do_modelo = modelo.summary()"
      ],
      "metadata": {
        "id": "No4U-UPDxzl9",
        "outputId": "8e69b995-5eea-4d89-b4f5-fd8afd47407c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_4\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " flatten_4 (Flatten)         (None, 3072)              0         \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 256)               786688    \n",
            "                                                                 \n",
            " dropout_4 (Dropout)         (None, 256)               0         \n",
            "                                                                 \n",
            " dense_9 (Dense)             (None, 10)                2570      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 789,258\n",
            "Trainable params: 789,258\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 9.0 Pesos da primeira camada DENSE - Processamento"
      ],
      "metadata": {
        "id": "TwuANSA2icf5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pesos_camada_dense = modelo.layers[1].get_weights()[0]\n",
        "pesos_camada_dense"
      ],
      "metadata": {
        "id": "gszwaw3VyHUz",
        "outputId": "4c4685d7-ab2d-48dc-e204-0e204726b8f8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 0.02792927, -0.00373939,  0.02032336, ...,  0.00971599,\n",
              "         0.00063605, -0.06021225],\n",
              "       [ 0.03567228, -0.05280499, -0.00902198, ...,  0.02659422,\n",
              "        -0.00750391, -0.02976294],\n",
              "       [ 0.02331997,  0.01466603, -0.02981322, ..., -0.0407236 ,\n",
              "         0.02940797, -0.05292826],\n",
              "       ...,\n",
              "       [-0.01128103, -0.05090598, -0.01610743, ..., -0.02693106,\n",
              "        -0.03179347,  0.04658198],\n",
              "       [-0.01129008, -0.02135096, -0.01395089, ..., -0.011921  ,\n",
              "         0.02133067, -0.01956587],\n",
              "       [ 0.03093136, -0.01011061, -0.04170971, ...,  0.00297252,\n",
              "        -0.00697012,  0.03805938]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 10.0 Vies na Camada de Processamento - DENSE"
      ],
      "metadata": {
        "id": "GPIMkDl3kU71"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "vieses_camada_dense = modelo.layers[1].get_weights()[1]\n",
        "vieses_camada_dense.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0kmz-ddpeHBU",
        "outputId": "fd306c84-fdbf-4e6d-ff24-297d42e1fa61"
      },
      "execution_count": 77,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(256,)"
            ]
          },
          "metadata": {},
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [],
      "metadata": {
        "id": "gO8MTnzen-IQ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 11 Configuração dos Pesos da camada DENSE"
      ],
      "metadata": {
        "id": "70M5y2cin-oI"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pesos_camada_dense.shape\n",
        "pesos_camada_dense_zerados = np.zeros((3072, 256))\n",
        "\n",
        "pesos_camada_dense_aleatorios = np.random.rand(3072, 256)\n",
        "\n",
        "modelo.layers[1].set_weights(\n",
        "    [pesos_camada_dense_aleatorios, vieses_camada_dense]\n",
        ")\n",
        "\n",
        "modelo.layers[1].get_weights()"
      ],
      "metadata": {
        "id": "wg6d-oO-yPa8",
        "outputId": "91709efb-1cf3-49f4-d818-5b9cbb542381",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([[0.25728577, 0.30497184, 0.661966  , ..., 0.77765423, 0.56938595,\n",
              "         0.06881054],\n",
              "        [0.34735072, 0.07548431, 0.8775067 , ..., 0.23060061, 0.6718545 ,\n",
              "         0.8325026 ],\n",
              "        [0.52872247, 0.18342587, 0.34149158, ..., 0.24032448, 0.67609227,\n",
              "         0.8024342 ],\n",
              "        ...,\n",
              "        [0.5705026 , 0.03690592, 0.87130123, ..., 0.7497586 , 0.12091012,\n",
              "         0.56237906],\n",
              "        [0.726661  , 0.86635876, 0.97118086, ..., 0.9831608 , 0.9889816 ,\n",
              "         0.27027482],\n",
              "        [0.46930966, 0.0354552 , 0.79516315, ..., 0.42381358, 0.01893902,\n",
              "         0.07200956]], dtype=float32),\n",
              " array([-0.00177983,  0.00490083, -0.00503956,  0.08778966,  0.24927288,\n",
              "        -0.00600217, -0.00600179, -0.00593143, -0.00651192, -0.00830725,\n",
              "        -0.00585748, -0.00552653, -0.00599461, -0.00717515, -0.0059841 ,\n",
              "        -0.00600412, -0.03029884, -0.00486894, -0.00275492, -0.00277288,\n",
              "        -0.00600469, -0.00441474, -0.00600466, -0.00286387, -0.00598939,\n",
              "        -0.00246761, -0.00600282, -0.00598886, -0.0043462 , -0.00679471,\n",
              "        -0.0064493 , -0.00496619, -0.02262495, -0.00644174, -0.00566366,\n",
              "         0.08013612, -0.1711026 , -0.00436232, -0.00510273, -0.00637011,\n",
              "        -0.00599483, -0.00598948, -0.00582642, -0.00610833, -0.00600456,\n",
              "        -0.00206918, -0.00528856, -0.00493828, -0.00514407, -0.00288341,\n",
              "        -0.00330492, -0.00557155, -0.00555663, -0.00594513, -0.00372839,\n",
              "         0.09442173, -0.00601003, -0.00613088, -0.04419461, -0.0044862 ,\n",
              "        -0.00600435, -0.00621067, -0.00600436, -0.00599972, -0.03826758,\n",
              "        -0.00374525, -0.00603931, -0.00564271, -0.0020413 , -0.00194668,\n",
              "        -0.00526205, -0.00583025, -0.00361226, -0.00449884, -0.00594855,\n",
              "        -0.00600168, -0.00539327, -0.00469251, -0.0063765 , -0.00500058,\n",
              "        -0.00779929,  0.        , -0.00663871, -0.00597891, -0.00608955,\n",
              "        -0.00600385, -0.08423001, -0.00489067, -0.00646218, -0.00600472,\n",
              "        -0.00600491, -0.00584063, -0.00528414, -0.00600398, -0.00439464,\n",
              "        -0.00785678, -0.00293918, -0.00600206, -0.00611847, -0.00600283,\n",
              "        -0.00601701, -0.00599101, -0.00319811, -0.0078978 , -0.00653568,\n",
              "        -0.00622585, -0.0013631 , -0.00674131, -0.00715098, -0.00598079,\n",
              "        -0.00597835, -0.0072668 , -0.00642691, -0.00530639, -0.00445001,\n",
              "        -0.0059853 , -0.00600488, -0.00655495, -0.00599751, -0.00510276,\n",
              "        -0.00358025, -0.00405048, -0.00677911, -0.00178175, -0.00311672,\n",
              "        -0.00618831, -0.00386064, -0.00485291, -0.04487478, -0.00604116,\n",
              "        -0.00417653, -0.00600243, -0.00598819, -0.00600319, -0.00600398,\n",
              "        -0.00527056, -0.00639609, -0.00676062, -0.00489461, -0.00495014,\n",
              "        -0.0054673 , -0.00527351, -0.00663157, -0.00355361, -0.00592715,\n",
              "        -0.00405748, -0.00600443, -0.00611046, -0.00499519, -0.00599144,\n",
              "        -0.00547869, -0.00600474, -0.00598178, -0.00396592, -0.00595759,\n",
              "        -0.0052636 , -0.00783657, -0.11388279, -0.0041121 , -0.00575867,\n",
              "        -0.00413607, -0.00447365, -0.00598984, -0.00206618, -0.00420425,\n",
              "        -0.00248836, -0.00028756, -0.00423837, -0.0029212 , -0.00582687,\n",
              "        -0.00337642, -0.07064746, -0.00572595, -0.00669784, -0.0060642 ,\n",
              "        -0.00147613, -0.00249326, -0.00191636, -0.005995  , -0.00600431,\n",
              "        -0.00626066, -0.00413959, -0.001866  , -0.00429457, -0.03237718,\n",
              "        -0.00366084, -0.00580962, -0.00421491, -0.0011801 , -0.00481214,\n",
              "        -0.00437633, -0.00438731, -0.00600628, -0.12287337, -0.00544733,\n",
              "        -0.00240907, -0.00593488, -0.00815526, -0.00426976, -0.00600161,\n",
              "        -0.00529949, -0.00226776, -0.00269026,  0.08888385, -0.00484188,\n",
              "        -0.00460505, -0.00474433, -0.00598659, -0.00152929, -0.00599911,\n",
              "         0.15111496, -0.00151572, -0.00607411, -0.00325098, -0.00457714,\n",
              "        -0.00599681, -0.00597614, -0.00600249, -0.00598086, -0.00598532,\n",
              "        -0.0036263 , -0.12339149, -0.00443338, -0.00605946, -0.00601695,\n",
              "        -0.00376562, -0.0059052 , -0.0060009 , -0.00426915, -0.00598313,\n",
              "        -0.00353126, -0.00734376, -0.21044268, -0.17464711, -0.00578335,\n",
              "        -0.00653324, -0.00582117, -0.07906517, -0.00509349, -0.00415593,\n",
              "        -0.00371449, -0.00580035, -0.0059974 , -0.00613201, -0.00596652,\n",
              "        -0.00577623, -0.00177619, -0.00578587, -0.00599169, -0.00578556,\n",
              "        -0.00524483,  0.04750657, -0.00466689, -0.00599199, -0.00600463,\n",
              "         0.21661104], dtype=float32)]"
            ]
          },
          "metadata": {},
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 12 Compilando Modelo\n"
      ],
      "metadata": {
        "id": "4PSlx1eJpfDg"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy',\n",
        "               metrics=['accuracy'])\n",
        "\n",
        "historico = modelo.fit(imagens_treino, identificacoes_treino, epochs=5, validation_split=0.2)"
      ],
      "metadata": {
        "id": "dqjj1Ic3F59_",
        "outputId": "9a7b660a-7cc6-42b6-db0d-e361482ec3f0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 79,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "1250/1250 [==============================] - 14s 11ms/step - loss: 470.2994 - accuracy: 0.1204 - val_loss: 52.0548 - val_accuracy: 0.2297\n",
            "Epoch 2/5\n",
            "1250/1250 [==============================] - 13s 10ms/step - loss: 98.5726 - accuracy: 0.1521 - val_loss: 37.9460 - val_accuracy: 0.1234\n",
            "Epoch 3/5\n",
            "1250/1250 [==============================] - 12s 10ms/step - loss: 28.3025 - accuracy: 0.1491 - val_loss: 14.8625 - val_accuracy: 0.1402\n",
            "Epoch 4/5\n",
            "1250/1250 [==============================] - 14s 11ms/step - loss: 15.9095 - accuracy: 0.1577 - val_loss: 8.5839 - val_accuracy: 0.1895\n",
            "Epoch 5/5\n",
            "1250/1250 [==============================] - 11s 9ms/step - loss: 7.0336 - accuracy: 0.1798 - val_loss: 3.2304 - val_accuracy: 0.2116\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 13 Plotando Grafico Acurácia X Épocas\n",
        "\n",
        "\n",
        "o Modelo apresenta problema de OVERFITTING"
      ],
      "metadata": {
        "id": "U52iDjHlpeVP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(historico.history['accuracy'])\n",
        "plt.plot(historico.history['val_accuracy'])\n",
        "plt.title('Acurácia por épocas')\n",
        "plt.xlabel('épocas')\n",
        "plt.ylabel('acurácia')\n",
        "plt.legend(['treino', 'validação'])\n"
      ],
      "metadata": {
        "id": "yP6n_O0wGhtI",
        "outputId": "189c0c10-3a4c-47f2-ec09-fee4065ade44",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        }
      },
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x79e3313617e0>"
            ]
          },
          "metadata": {},
          "execution_count": 80
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 14 Plotando Grafico Perdas X ÉpocasNova seção"
      ],
      "metadata": {
        "id": "FJuh2vUIrcLX"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(historico.history['loss'])\n",
        "plt.plot(historico.history['val_loss'])\n",
        "plt.title('Perda por épocas')\n",
        "plt.xlabel('épocas')\n",
        "plt.ylabel('perda')\n",
        "plt.legend(['treino', 'validação'])"
      ],
      "metadata": {
        "id": "8fIFMVuuHOCQ",
        "outputId": "dbaa1503-22bc-4fa4-d1e0-06dde3ea1111",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 490
        }
      },
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x79e31ddfda50>"
            ]
          },
          "metadata": {},
          "execution_count": 82
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "metadata": {
        "id": "4-RqzS88dgk_"
      },
      "cell_type": "markdown",
      "source": [
        "# 15 Salvando e carregando o modelo treinado"
      ]
    },
    {
      "metadata": {
        "id": "SVjmU41IELzX"
      },
      "cell_type": "code",
      "source": [
        "modelo.save('modelo.h5')\n",
        "modelo_salvo = load_model('modelo.h5')"
      ],
      "execution_count": 83,
      "outputs": []
    },
    {
      "metadata": {
        "id": "JFXG1oibd8E8"
      },
      "cell_type": "markdown",
      "source": [
        "# 16 Testando o modelo salvo"
      ]
    },
    {
      "metadata": {
        "id": "I2kLEHPYUdhd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "bda3b242-f1d1-4359-8fb4-a12efc39279b"
      },
      "cell_type": "code",
      "source": [
        "testes = modelo.predict(imagens_teste)\n",
        "print('resultado teste:', np.argmax(testes[1]))\n",
        "print('número da imagem de teste:', identificacoes_teste[1])\n",
        "\n",
        "testes_modelo_salvo = modelo_salvo.predict(imagens_teste)\n",
        "print('resultado teste modelo salvo:', np.argmax(testes_modelo_salvo[1]))\n",
        "print('número da imagem de teste:', identificacoes_teste[1])"
      ],
      "execution_count": 84,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 3ms/step\n",
            "resultado teste: 9\n",
            "número da imagem de teste: [8]\n",
            "313/313 [==============================] - 1s 4ms/step\n",
            "resultado teste modelo salvo: 9\n",
            "número da imagem de teste: [8]\n"
          ]
        }
      ]
    },
    {
      "metadata": {
        "id": "U64QMV1GeCUr"
      },
      "cell_type": "markdown",
      "source": [
        "# 17 Avaliando o modelo"
      ]
    },
    {
      "metadata": {
        "id": "wsbr9WqyXQ2G",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0fe8b2a0-c19e-484f-c7e3-d45f79aa3009"
      },
      "cell_type": "code",
      "source": [
        "perda_teste, acuracia_teste = modelo.evaluate(imagens_teste, identificacoes_teste)\n",
        "print('Perda do teste:', perda_teste)\n",
        "print('Acurácia do teste:', acuracia_teste)"
      ],
      "execution_count": 85,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 4ms/step - loss: 673.2881 - accuracy: 0.1979\n",
            "Perda do teste: 673.2881469726562\n",
            "Acurácia do teste: 0.19789999723434448\n"
          ]
        }
      ]
    }
  ]
}