Kaggle_Titanic.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Kaggle Titanic.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Maratinhos/hw/blob/master/Kaggle_Titanic.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DAg1NfO6w0Hy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Conv2D, MaxPooling2D, Dropout, Flatten, BatchNormalization\n",
        "from tensorflow.keras import utils\n",
        "from tensorflow.keras.preprocessing import image\n",
        "from tensorflow.python.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.callbacks import ReduceLROnPlateau, ModelCheckpoint, EarlyStopping\n",
        "import tensorflow as tf\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from PIL import Image\n",
        "from google.colab import files\n",
        "import pandas as pd\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HoCxiBqgxDoL",
        "colab_type": "code",
        "outputId": "31244c62-4afa-4eca-c2da-0657c0c2d9e4",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 74
        }
      },
      "source": [
        "file = files.upload()"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-f219ee54-3e64-49d9-a575-3b8ff0d9f76d\" name=\"files[]\" multiple disabled />\n",
              "     <output id=\"result-f219ee54-3e64-49d9-a575-3b8ff0d9f76d\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving kaggle.json to kaggle.json\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YrSIN64cxJUn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!mkdir ~/.kaggle\n",
        "!mv kaggle.json ~/.kaggle\n",
        "!chmod 600 /root/.kaggle/kaggle.json\n",
        "!kaggle competitions download -c titanic"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wyQdM9t0xQJ6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!kaggle competitions download -c titanic"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LIVtWtigxdex",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df = pd.read_csv('train.csv')\n",
        "df_test = pd.read_csv('test.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o4d4PtdvyHfo",
        "colab_type": "code",
        "outputId": "ed81e1d7-2649-4950-a2ae-0062ed7e4018",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>PassengerId</th>\n",
              "      <th>Survived</th>\n",
              "      <th>Pclass</th>\n",
              "      <th>Name</th>\n",
              "      <th>Sex</th>\n",
              "      <th>Age</th>\n",
              "      <th>SibSp</th>\n",
              "      <th>Parch</th>\n",
              "      <th>Ticket</th>\n",
              "      <th>Fare</th>\n",
              "      <th>Cabin</th>\n",
              "      <th>Embarked</th>\n",
              "      <th>Cabin2</th>\n",
              "      <th>rod</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>Braund, Mr. Owen Harris</td>\n",
              "      <td>male</td>\n",
              "      <td>22.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>A/5 21171</td>\n",
              "      <td>7.2500</td>\n",
              "      <td>NaN</td>\n",
              "      <td>S</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
              "      <td>female</td>\n",
              "      <td>38.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>PC 17599</td>\n",
              "      <td>71.2833</td>\n",
              "      <td>C85</td>\n",
              "      <td>C</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>Heikkinen, Miss. Laina</td>\n",
              "      <td>female</td>\n",
              "      <td>26.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>STON/O2. 3101282</td>\n",
              "      <td>7.9250</td>\n",
              "      <td>NaN</td>\n",
              "      <td>S</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
              "      <td>female</td>\n",
              "      <td>35.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>113803</td>\n",
              "      <td>53.1000</td>\n",
              "      <td>C123</td>\n",
              "      <td>S</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>Allen, Mr. William Henry</td>\n",
              "      <td>male</td>\n",
              "      <td>35.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>373450</td>\n",
              "      <td>8.0500</td>\n",
              "      <td>NaN</td>\n",
              "      <td>S</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   PassengerId  Survived  Pclass  ... Embarked Cabin2  rod\n",
              "0            1         0       3  ...        S      1    1\n",
              "1            2         1       1  ...        C      0    1\n",
              "2            3         1       3  ...        S      1    0\n",
              "3            4         1       1  ...        S      0    1\n",
              "4            5         0       3  ...        S      1    0\n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1HXfKS1X2Fpr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df['Sex'] = df['Sex'].replace('male', 0)\n",
        "df['Sex'] = df['Sex'].replace('female', 1)\n",
        "df['Embarked'] = df['Embarked'].replace('S', 0)\n",
        "df['Embarked'] = df['Embarked'].replace('C', 1)\n",
        "df['Embarked'] = df['Embarked'].replace('Q', 2)\n",
        "df['Cabin2'] = pd.isnull(df['Cabin'])\n",
        "df['Cabin2'] = df['Cabin2'].astype(int)\n",
        "df['rod'] = df['SibSp'] + df['Parch']\n",
        "\n",
        "df_test['Sex'] = df_test['Sex'].replace('male', 0)\n",
        "df_test['Sex'] = df_test['Sex'].replace('female', 1)\n",
        "df_test['Embarked'] = df_test['Embarked'].replace('S', 0)\n",
        "df_test['Embarked'] = df_test['Embarked'].replace('C', 1)\n",
        "df_test['Embarked'] = df_test['Embarked'].replace('Q', 2)\n",
        "df_test['Cabin2'] = pd.isnull(df_test['Cabin'])\n",
        "df_test['Cabin2'] = df_test['Cabin2'].astype(int)\n",
        "df_test['rod'] = df_test['SibSp'] + df_test['Parch']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s3pfXI678m9M",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df2 = df[['Pclass', 'Sex', 'Age', 'rod', 'Fare', 'Embarked', 'Cabin2', 'Survived']]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n6rMzvPi9egK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df3 = df2.dropna()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zbNE5-0LGQGr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x = df3[['Pclass', 'Sex', 'Age', 'rod', 'Fare', 'Embarked', 'Cabin2']]\n",
        "y = df3['Survived']\n",
        "\n",
        "x_test = df_test[['Pclass', 'Sex', 'Age', 'rod', 'Fare', 'Embarked', 'Cabin2']]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Clb3M_PnfMzl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x = x.fillna(x.mean())\n",
        "\n",
        "x_test = x_test.fillna(x_test.mean())"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5V7Y7_sNHhSk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train, x_val, y_train, y_val = train_test_split(x,y,test_size=0.1, shuffle=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "imdIytGqI08o",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "scaler = StandardScaler()\n",
        "x_train_scaled = scaler.fit_transform(x_train)\n",
        "x_val_scaled = scaler.fit_transform(x_val)\n",
        "\n",
        "x_test_scaled = scaler.fit_transform(x_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fw9tsUN9Os8q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train_scaled = x_train_scaled.astype(np.float32)\n",
        "x_val_scaled = x_val_scaled.astype(np.float32)\n",
        "y_train = y_train.astype(np.float32)\n",
        "y_val = y_val.astype(np.float32)\n",
        "\n",
        "x_test_scaled = x_test_scaled.astype(np.float32)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p-6u3a1qUcZP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train_scaled = np.array(x_train_scaled)\n",
        "x_val_scaled = np.array(x_val_scaled)\n",
        "y_train = np.array(y_train)\n",
        "y_val = np.array(y_val)\n",
        "\n",
        "x_test_scaled = np.array(x_test_scaled)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hwGnacrkL4Al",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = Sequential([\n",
        "    Dense(16, activation='relu', input_shape=(7,)),\n",
        "    Dense(8, activation='relu'),\n",
        "    Dense(4, activation='relu'),\n",
        "    Dropout(0.25),\n",
        "    Dense(2, activation='relu'),\n",
        "    #BatchNormalization(),\n",
        "    #Dropout(0.2),\n",
        "    #Dense(1024, activation='relu'),\n",
        "    #BatchNormalization(),\n",
        "    #Dropout(0.01),\n",
        "    #Dense(1024, activation='relu'),\n",
        "    #BatchNormalization(),\n",
        "    #Dropout(0.01),\n",
        "    #Dense(512, activation='relu'),\n",
        "    #BatchNormalization(),\n",
        "    #Dropout(0.01),\n",
        "    Dense(1, activation='sigmoid')\n",
        "])\n",
        "\n",
        "model.compile(optimizer = 'adam', \n",
        "              loss='binary_crossentropy', \n",
        "              metrics = ['accuracy'])\n",
        "\n",
        "checkpoint = ModelCheckpoint('kaggle-titanic.hdf5', monitor = 'val_acc', save_best_only = True, verbose = 1, mode = 'max')\n",
        "learning_rate_reduction = ReduceLROnPlateau(monitor = 'val_acc', patience = 3, verbose = 1, factor = 0.3, min_lr = 0.1)\n",
        "early_stopping = EarlyStopping(monitor = 'val_loss', patience = 10, verbose = 1, restore_best_weights = True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4thyQ3gvNaJ7",
        "colab_type": "code",
        "outputId": "7d79948b-453e-4252-c0cc-9a7d1f5279c7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history = model.fit(x_train_scaled, \n",
        "                    y_train, \n",
        "                    batch_size = 128,\n",
        "                    validation_data = (x_val_scaled, y_val), \n",
        "                    epochs = 100,\n",
        "                    verbose = 1\n",
        "                    #callbacks = [checkpoint, learning_rate_reduction, early_stopping]\n",
        "                   )"
      ],
      "execution_count": 131,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 640 samples, validate on 72 samples\n",
            "Epoch 1/100\n",
            "640/640 [==============================] - 2s 3ms/sample - loss: 0.6966 - acc: 0.4016 - val_loss: 0.6955 - val_acc: 0.3056\n",
            "Epoch 2/100\n",
            "640/640 [==============================] - 0s 58us/sample - loss: 0.6949 - acc: 0.4266 - val_loss: 0.6938 - val_acc: 0.3889\n",
            "Epoch 3/100\n",
            "640/640 [==============================] - 0s 51us/sample - loss: 0.6932 - acc: 0.4797 - val_loss: 0.6924 - val_acc: 0.6667\n",
            "Epoch 4/100\n",
            "640/640 [==============================] - 0s 54us/sample - loss: 0.6917 - acc: 0.5672 - val_loss: 0.6912 - val_acc: 0.6250\n",
            "Epoch 5/100\n",
            "640/640 [==============================] - 0s 58us/sample - loss: 0.6904 - acc: 0.6156 - val_loss: 0.6903 - val_acc: 0.6111\n",
            "Epoch 6/100\n",
            "640/640 [==============================] - 0s 53us/sample - loss: 0.6887 - acc: 0.6609 - val_loss: 0.6896 - val_acc: 0.6111\n",
            "Epoch 7/100\n",
            "640/640 [==============================] - 0s 61us/sample - loss: 0.6883 - acc: 0.6484 - val_loss: 0.6888 - val_acc: 0.6250\n",
            "Epoch 8/100\n",
            "640/640 [==============================] - 0s 56us/sample - loss: 0.6866 - acc: 0.6609 - val_loss: 0.6876 - val_acc: 0.6250\n",
            "Epoch 9/100\n",
            "640/640 [==============================] - 0s 49us/sample - loss: 0.6860 - acc: 0.6656 - val_loss: 0.6861 - val_acc: 0.6389\n",
            "Epoch 10/100\n",
            "640/640 [==============================] - 0s 61us/sample - loss: 0.6848 - acc: 0.6719 - val_loss: 0.6842 - val_acc: 0.6528\n",
            "Epoch 11/100\n",
            "640/640 [==============================] - 0s 65us/sample - loss: 0.6837 - acc: 0.6703 - val_loss: 0.6820 - val_acc: 0.7222\n",
            "Epoch 12/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.6817 - acc: 0.6891 - val_loss: 0.6793 - val_acc: 0.7222\n",
            "Epoch 13/100\n",
            "640/640 [==============================] - 0s 54us/sample - loss: 0.6808 - acc: 0.6797 - val_loss: 0.6762 - val_acc: 0.7361\n",
            "Epoch 14/100\n",
            "640/640 [==============================] - 0s 50us/sample - loss: 0.6785 - acc: 0.6906 - val_loss: 0.6728 - val_acc: 0.7500\n",
            "Epoch 15/100\n",
            "640/640 [==============================] - 0s 65us/sample - loss: 0.6747 - acc: 0.7125 - val_loss: 0.6692 - val_acc: 0.7917\n",
            "Epoch 16/100\n",
            "640/640 [==============================] - 0s 58us/sample - loss: 0.6732 - acc: 0.7125 - val_loss: 0.6648 - val_acc: 0.7917\n",
            "Epoch 17/100\n",
            "640/640 [==============================] - 0s 56us/sample - loss: 0.6694 - acc: 0.7297 - val_loss: 0.6600 - val_acc: 0.8056\n",
            "Epoch 18/100\n",
            "640/640 [==============================] - 0s 59us/sample - loss: 0.6654 - acc: 0.7312 - val_loss: 0.6549 - val_acc: 0.8194\n",
            "Epoch 19/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.6670 - acc: 0.7000 - val_loss: 0.6497 - val_acc: 0.8194\n",
            "Epoch 20/100\n",
            "640/640 [==============================] - 0s 55us/sample - loss: 0.6591 - acc: 0.7266 - val_loss: 0.6443 - val_acc: 0.8333\n",
            "Epoch 21/100\n",
            "640/640 [==============================] - 0s 55us/sample - loss: 0.6591 - acc: 0.7109 - val_loss: 0.6383 - val_acc: 0.8472\n",
            "Epoch 22/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.6556 - acc: 0.7188 - val_loss: 0.6320 - val_acc: 0.8472\n",
            "Epoch 23/100\n",
            "640/640 [==============================] - 0s 53us/sample - loss: 0.6520 - acc: 0.7297 - val_loss: 0.6253 - val_acc: 0.8472\n",
            "Epoch 24/100\n",
            "640/640 [==============================] - 0s 58us/sample - loss: 0.6466 - acc: 0.7203 - val_loss: 0.6181 - val_acc: 0.8472\n",
            "Epoch 25/100\n",
            "640/640 [==============================] - 0s 64us/sample - loss: 0.6458 - acc: 0.7094 - val_loss: 0.6116 - val_acc: 0.8472\n",
            "Epoch 26/100\n",
            "640/640 [==============================] - 0s 63us/sample - loss: 0.6352 - acc: 0.7344 - val_loss: 0.6044 - val_acc: 0.8611\n",
            "Epoch 27/100\n",
            "640/640 [==============================] - 0s 55us/sample - loss: 0.6344 - acc: 0.7391 - val_loss: 0.5970 - val_acc: 0.8472\n",
            "Epoch 28/100\n",
            "640/640 [==============================] - 0s 58us/sample - loss: 0.6384 - acc: 0.7109 - val_loss: 0.5896 - val_acc: 0.8472\n",
            "Epoch 29/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.6323 - acc: 0.7250 - val_loss: 0.5835 - val_acc: 0.8472\n",
            "Epoch 30/100\n",
            "640/640 [==============================] - 0s 63us/sample - loss: 0.6185 - acc: 0.7500 - val_loss: 0.5767 - val_acc: 0.8333\n",
            "Epoch 31/100\n",
            "640/640 [==============================] - 0s 63us/sample - loss: 0.6182 - acc: 0.7437 - val_loss: 0.5704 - val_acc: 0.8611\n",
            "Epoch 32/100\n",
            "640/640 [==============================] - 0s 61us/sample - loss: 0.6122 - acc: 0.7406 - val_loss: 0.5636 - val_acc: 0.8472\n",
            "Epoch 33/100\n",
            "640/640 [==============================] - 0s 58us/sample - loss: 0.6139 - acc: 0.7406 - val_loss: 0.5573 - val_acc: 0.8472\n",
            "Epoch 34/100\n",
            "640/640 [==============================] - 0s 64us/sample - loss: 0.6096 - acc: 0.7406 - val_loss: 0.5522 - val_acc: 0.8472\n",
            "Epoch 35/100\n",
            "640/640 [==============================] - 0s 61us/sample - loss: 0.6024 - acc: 0.7563 - val_loss: 0.5461 - val_acc: 0.8472\n",
            "Epoch 36/100\n",
            "640/640 [==============================] - 0s 60us/sample - loss: 0.6131 - acc: 0.7391 - val_loss: 0.5409 - val_acc: 0.8472\n",
            "Epoch 37/100\n",
            "640/640 [==============================] - 0s 55us/sample - loss: 0.5962 - acc: 0.7625 - val_loss: 0.5363 - val_acc: 0.8611\n",
            "Epoch 38/100\n",
            "640/640 [==============================] - 0s 60us/sample - loss: 0.6056 - acc: 0.7453 - val_loss: 0.5314 - val_acc: 0.8611\n",
            "Epoch 39/100\n",
            "640/640 [==============================] - 0s 55us/sample - loss: 0.6049 - acc: 0.7328 - val_loss: 0.5279 - val_acc: 0.8750\n",
            "Epoch 40/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.6006 - acc: 0.7500 - val_loss: 0.5250 - val_acc: 0.8750\n",
            "Epoch 41/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.6078 - acc: 0.7344 - val_loss: 0.5226 - val_acc: 0.8750\n",
            "Epoch 42/100\n",
            "640/640 [==============================] - 0s 54us/sample - loss: 0.6005 - acc: 0.7516 - val_loss: 0.5206 - val_acc: 0.8889\n",
            "Epoch 43/100\n",
            "640/640 [==============================] - 0s 56us/sample - loss: 0.5934 - acc: 0.7531 - val_loss: 0.5173 - val_acc: 0.8889\n",
            "Epoch 44/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.5973 - acc: 0.7500 - val_loss: 0.5148 - val_acc: 0.8889\n",
            "Epoch 45/100\n",
            "640/640 [==============================] - 0s 49us/sample - loss: 0.5931 - acc: 0.7500 - val_loss: 0.5122 - val_acc: 0.8889\n",
            "Epoch 46/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5826 - acc: 0.7734 - val_loss: 0.5087 - val_acc: 0.8889\n",
            "Epoch 47/100\n",
            "640/640 [==============================] - 0s 53us/sample - loss: 0.5857 - acc: 0.7609 - val_loss: 0.5059 - val_acc: 0.8889\n",
            "Epoch 48/100\n",
            "640/640 [==============================] - 0s 49us/sample - loss: 0.5857 - acc: 0.7625 - val_loss: 0.5026 - val_acc: 0.8889\n",
            "Epoch 49/100\n",
            "640/640 [==============================] - 0s 55us/sample - loss: 0.5886 - acc: 0.7516 - val_loss: 0.5006 - val_acc: 0.8889\n",
            "Epoch 50/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5861 - acc: 0.7563 - val_loss: 0.5001 - val_acc: 0.8889\n",
            "Epoch 51/100\n",
            "640/640 [==============================] - 0s 51us/sample - loss: 0.5755 - acc: 0.7703 - val_loss: 0.4979 - val_acc: 0.8750\n",
            "Epoch 52/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5794 - acc: 0.7719 - val_loss: 0.4941 - val_acc: 0.8750\n",
            "Epoch 53/100\n",
            "640/640 [==============================] - 0s 51us/sample - loss: 0.5815 - acc: 0.7594 - val_loss: 0.4916 - val_acc: 0.8750\n",
            "Epoch 54/100\n",
            "640/640 [==============================] - 0s 49us/sample - loss: 0.5780 - acc: 0.7609 - val_loss: 0.4899 - val_acc: 0.8750\n",
            "Epoch 55/100\n",
            "640/640 [==============================] - 0s 46us/sample - loss: 0.5996 - acc: 0.7297 - val_loss: 0.4895 - val_acc: 0.8750\n",
            "Epoch 56/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.5794 - acc: 0.7625 - val_loss: 0.4889 - val_acc: 0.8750\n",
            "Epoch 57/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.5788 - acc: 0.7594 - val_loss: 0.4873 - val_acc: 0.8750\n",
            "Epoch 58/100\n",
            "640/640 [==============================] - 0s 51us/sample - loss: 0.5678 - acc: 0.7672 - val_loss: 0.4854 - val_acc: 0.8750\n",
            "Epoch 59/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.5863 - acc: 0.7516 - val_loss: 0.4824 - val_acc: 0.8750\n",
            "Epoch 60/100\n",
            "640/640 [==============================] - 0s 64us/sample - loss: 0.5918 - acc: 0.7328 - val_loss: 0.4821 - val_acc: 0.8889\n",
            "Epoch 61/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.5719 - acc: 0.7578 - val_loss: 0.4805 - val_acc: 0.8889\n",
            "Epoch 62/100\n",
            "640/640 [==============================] - 0s 59us/sample - loss: 0.5768 - acc: 0.7469 - val_loss: 0.4790 - val_acc: 0.8750\n",
            "Epoch 63/100\n",
            "640/640 [==============================] - 0s 45us/sample - loss: 0.5764 - acc: 0.7609 - val_loss: 0.4777 - val_acc: 0.8750\n",
            "Epoch 64/100\n",
            "640/640 [==============================] - 0s 47us/sample - loss: 0.5663 - acc: 0.7625 - val_loss: 0.4764 - val_acc: 0.9028\n",
            "Epoch 65/100\n",
            "640/640 [==============================] - 0s 50us/sample - loss: 0.5773 - acc: 0.7391 - val_loss: 0.4752 - val_acc: 0.8889\n",
            "Epoch 66/100\n",
            "640/640 [==============================] - 0s 50us/sample - loss: 0.5693 - acc: 0.7516 - val_loss: 0.4731 - val_acc: 0.9028\n",
            "Epoch 67/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.5626 - acc: 0.7719 - val_loss: 0.4720 - val_acc: 0.9028\n",
            "Epoch 68/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.5655 - acc: 0.7672 - val_loss: 0.4688 - val_acc: 0.9028\n",
            "Epoch 69/100\n",
            "640/640 [==============================] - 0s 49us/sample - loss: 0.5650 - acc: 0.7656 - val_loss: 0.4667 - val_acc: 0.9028\n",
            "Epoch 70/100\n",
            "640/640 [==============================] - 0s 54us/sample - loss: 0.5721 - acc: 0.7594 - val_loss: 0.4647 - val_acc: 0.9028\n",
            "Epoch 71/100\n",
            "640/640 [==============================] - 0s 54us/sample - loss: 0.5655 - acc: 0.7578 - val_loss: 0.4638 - val_acc: 0.9028\n",
            "Epoch 72/100\n",
            "640/640 [==============================] - 0s 44us/sample - loss: 0.5609 - acc: 0.7672 - val_loss: 0.4640 - val_acc: 0.8889\n",
            "Epoch 73/100\n",
            "640/640 [==============================] - 0s 53us/sample - loss: 0.5655 - acc: 0.7656 - val_loss: 0.4646 - val_acc: 0.8889\n",
            "Epoch 74/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5491 - acc: 0.7828 - val_loss: 0.4616 - val_acc: 0.8889\n",
            "Epoch 75/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5589 - acc: 0.7719 - val_loss: 0.4589 - val_acc: 0.8889\n",
            "Epoch 76/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.5498 - acc: 0.7797 - val_loss: 0.4574 - val_acc: 0.8889\n",
            "Epoch 77/100\n",
            "640/640 [==============================] - 0s 59us/sample - loss: 0.5645 - acc: 0.7594 - val_loss: 0.4553 - val_acc: 0.8889\n",
            "Epoch 78/100\n",
            "640/640 [==============================] - 0s 69us/sample - loss: 0.5568 - acc: 0.7625 - val_loss: 0.4553 - val_acc: 0.8889\n",
            "Epoch 79/100\n",
            "640/640 [==============================] - 0s 46us/sample - loss: 0.5625 - acc: 0.7625 - val_loss: 0.4550 - val_acc: 0.8889\n",
            "Epoch 80/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5648 - acc: 0.7594 - val_loss: 0.4552 - val_acc: 0.8889\n",
            "Epoch 81/100\n",
            "640/640 [==============================] - 0s 44us/sample - loss: 0.5449 - acc: 0.7844 - val_loss: 0.4539 - val_acc: 0.8889\n",
            "Epoch 82/100\n",
            "640/640 [==============================] - 0s 49us/sample - loss: 0.5458 - acc: 0.7766 - val_loss: 0.4513 - val_acc: 0.8889\n",
            "Epoch 83/100\n",
            "640/640 [==============================] - 0s 54us/sample - loss: 0.5526 - acc: 0.7797 - val_loss: 0.4502 - val_acc: 0.8889\n",
            "Epoch 84/100\n",
            "640/640 [==============================] - 0s 50us/sample - loss: 0.5531 - acc: 0.7641 - val_loss: 0.4478 - val_acc: 0.8889\n",
            "Epoch 85/100\n",
            "640/640 [==============================] - 0s 54us/sample - loss: 0.5455 - acc: 0.7828 - val_loss: 0.4467 - val_acc: 0.8889\n",
            "Epoch 86/100\n",
            "640/640 [==============================] - 0s 65us/sample - loss: 0.5449 - acc: 0.7781 - val_loss: 0.4453 - val_acc: 0.8889\n",
            "Epoch 87/100\n",
            "640/640 [==============================] - 0s 48us/sample - loss: 0.5453 - acc: 0.7766 - val_loss: 0.4438 - val_acc: 0.8889\n",
            "Epoch 88/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.5426 - acc: 0.7781 - val_loss: 0.4429 - val_acc: 0.8889\n",
            "Epoch 89/100\n",
            "640/640 [==============================] - 0s 50us/sample - loss: 0.5413 - acc: 0.7812 - val_loss: 0.4427 - val_acc: 0.8889\n",
            "Epoch 90/100\n",
            "640/640 [==============================] - 0s 57us/sample - loss: 0.5515 - acc: 0.7578 - val_loss: 0.4424 - val_acc: 0.8889\n",
            "Epoch 91/100\n",
            "640/640 [==============================] - 0s 55us/sample - loss: 0.5380 - acc: 0.7750 - val_loss: 0.4406 - val_acc: 0.8889\n",
            "Epoch 92/100\n",
            "640/640 [==============================] - 0s 58us/sample - loss: 0.5424 - acc: 0.7641 - val_loss: 0.4390 - val_acc: 0.8889\n",
            "Epoch 93/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5523 - acc: 0.7703 - val_loss: 0.4374 - val_acc: 0.8889\n",
            "Epoch 94/100\n",
            "640/640 [==============================] - 0s 51us/sample - loss: 0.5345 - acc: 0.7797 - val_loss: 0.4361 - val_acc: 0.8889\n",
            "Epoch 95/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5276 - acc: 0.7937 - val_loss: 0.4344 - val_acc: 0.8889\n",
            "Epoch 96/100\n",
            "640/640 [==============================] - 0s 50us/sample - loss: 0.5389 - acc: 0.7812 - val_loss: 0.4335 - val_acc: 0.8889\n",
            "Epoch 97/100\n",
            "640/640 [==============================] - 0s 51us/sample - loss: 0.5400 - acc: 0.7797 - val_loss: 0.4325 - val_acc: 0.8889\n",
            "Epoch 98/100\n",
            "640/640 [==============================] - 0s 59us/sample - loss: 0.5274 - acc: 0.7828 - val_loss: 0.4298 - val_acc: 0.8889\n",
            "Epoch 99/100\n",
            "640/640 [==============================] - 0s 51us/sample - loss: 0.5239 - acc: 0.7875 - val_loss: 0.4272 - val_acc: 0.8889\n",
            "Epoch 100/100\n",
            "640/640 [==============================] - 0s 52us/sample - loss: 0.5390 - acc: 0.7719 - val_loss: 0.4260 - val_acc: 0.8889\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kcV35eqRjIJr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.load_weights('kaggle-titanic.hdf5')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NCV2_n74NmBC",
        "colab_type": "code",
        "outputId": "88d06c31-1db6-4255-a14b-2edab4a36ecb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        }
      },
      "source": [
        "# Training plots\n",
        "epochs = [i for i in range(1, len(history.history['loss'])+1)]\n",
        "\n",
        "plt.plot(epochs, history.history['loss'], color='blue', label=\"training_loss\")\n",
        "plt.plot(epochs, history.history['val_loss'], color='red', label=\"validation_loss\")\n",
        "plt.legend(loc='best')\n",
        "plt.title('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.show()\n",
        "\n",
        "plt.plot(epochs, history.history['acc'], color='blue', label=\"training_accuracy\")\n",
        "plt.plot(epochs, history.history['val_acc'], color='red',label=\"validation_accuracy\")\n",
        "plt.legend(loc='best')\n",
        "plt.title('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.show()"
      ],
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEWCAYAAACKSkfIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxgQRC/rGGBNELOgbY0wQ+X+wasPRd8nZDgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEWCAYAAACdaNcBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5lNElYr9ivPm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred = model.predict(x_test_scaled)\n",
        "y_final = (y_pred > 0.5).astype(int).reshape(df_test.shape[0])\n",
        "    \n",
        "output = pd.DataFrame({'PassengerId': df_test['PassengerId'], 'Survived': y_final})\n",
        "output.to_csv('submission.csv', index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vRZOKgUbsiGM",
        "colab_type": "code",
        "outputId": "8ac90041-dece-48e4-82b7-f06eed9d50f5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "output.head()"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>PassengerId</th>\n",
              "      <th>Survived</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>892</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>893</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>894</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>895</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>896</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   PassengerId  Survived\n",
              "0          892         0\n",
              "1          893         0\n",
              "2          894         0\n",
              "3          895         0\n",
              "4          896         0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "410ki_vSmL4q",
        "colab_type": "code",
        "outputId": "50b39879-f3bf-4811-b135-d411fef82a46",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        }
      },
      "source": [
        "output.describe()"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>PassengerId</th>\n",
              "      <th>Survived</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>418.000000</td>\n",
              "      <td>418.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>1100.500000</td>\n",
              "      <td>0.322967</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>120.810458</td>\n",
              "      <td>0.468170</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>892.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>996.250000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>1100.500000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>1204.750000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1309.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       PassengerId    Survived\n",
              "count   418.000000  418.000000\n",
              "mean   1100.500000    0.322967\n",
              "std     120.810458    0.468170\n",
              "min     892.000000    0.000000\n",
              "25%     996.250000    0.000000\n",
              "50%    1100.500000    0.000000\n",
              "75%    1204.750000    1.000000\n",
              "max    1309.000000    1.000000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V9F4bBCSmOqv",
        "colab_type": "code",
        "outputId": "d0ef5bdf-5d08-4f05-a9ce-46d2044ca9e6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "!kaggle competitions submit -c titanic -f submission.csv -m \"sub #5\""
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "100% 2.77k/2.77k [00:10<00:00, 270B/s]\n",
            "Successfully submitted to Titanic: Machine Learning from Disaster"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}