woche2/notebooks/first-mnist-net/mnist-commented-solution.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "mnist-commented-solution.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4Pb7pbcw_mu3",
        "colab_type": "text"
      },
      "source": [
        "#### Importieren der Libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BYKdhdArMixR",
        "colab_type": "code",
        "outputId": "445bec43-d8e4-4ac0-afa6-172255169afe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 386
        }
      },
      "source": [
        "!pip install deeplearning2020\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from deeplearning2020 import helpers"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting deeplearning2020\n",
            "  Downloading https://files.pythonhosted.org/packages/fe/33/4cb24b085b1dc2a333d7553cc3135bb6ea95384fc685ffc4c17bab41a609/deeplearning2020-0.2.0.tar.gz\n",
            "Collecting kerasltisubmission>=0.3.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/7e/7b/810826936ce9ff3b05f9b4e2f857a289581b7cabb96d88059194c459ffb1/kerasltisubmission-0.3.0.tar.gz (390kB)\n",
            "\u001b[K     |████████████████████████████████| 399kB 6.9MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from kerasltisubmission>=0.3.0->deeplearning2020) (1.17.5)\n",
            "Requirement already satisfied: progressbar2 in /usr/local/lib/python3.6/dist-packages (from kerasltisubmission>=0.3.0->deeplearning2020) (3.38.0)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from progressbar2->kerasltisubmission>=0.3.0->deeplearning2020) (1.12.0)\n",
            "Requirement already satisfied: python-utils>=2.3.0 in /usr/local/lib/python3.6/dist-packages (from progressbar2->kerasltisubmission>=0.3.0->deeplearning2020) (2.3.0)\n",
            "Building wheels for collected packages: deeplearning2020, kerasltisubmission\n",
            "  Building wheel for deeplearning2020 (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for deeplearning2020: filename=deeplearning2020-0.2.0-py2.py3-none-any.whl size=5954 sha256=26e3974607aeaf4db63d72f6959ee467b62ae2fc8aa127e22e40b814994bae23\n",
            "  Stored in directory: /root/.cache/pip/wheels/d9/62/f9/503c2046ee773650f8c0adf0c068e9caff74755f396e51e200\n",
            "  Building wheel for kerasltisubmission (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for kerasltisubmission: filename=kerasltisubmission-0.3.0-py2.py3-none-any.whl size=6545 sha256=3e4eb4b031cb080b98cf0f593d833dda5c10a5b5a5ba38b742d65dd095e52392\n",
            "  Stored in directory: /root/.cache/pip/wheels/93/36/c1/184c1a6e1434cdc13b829d82c9de948f042d846b88c67646c8\n",
            "Successfully built deeplearning2020 kerasltisubmission\n",
            "Installing collected packages: kerasltisubmission, deeplearning2020\n",
            "Successfully installed deeplearning2020-0.2.0 kerasltisubmission-0.3.0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<p style=\"color: red;\">\n",
              "The default version of TensorFlow in Colab will soon switch to TensorFlow 2.x.<br>\n",
              "We recommend you <a href=\"https://www.tensorflow.org/guide/migrate\" target=\"_blank\">upgrade</a> now \n",
              "or ensure your notebook will continue to use TensorFlow 1.x via the <code>%tensorflow_version 1.x</code> magic:\n",
              "<a href=\"https://colab.research.google.com/notebooks/tensorflow_version.ipynb\" target=\"_blank\">more info</a>.</p>\n"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h4yynkUtAh4U",
        "colab_type": "text"
      },
      "source": [
        "#### Laden des MNIST Dataset\n",
        "\n",
        "Als Erstes wollen wir den Datensatz wie im Video \"Laden und Bearbeiten des MNIST Datensatz\" laden."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HrkKkPW2OIfB",
        "colab_type": "code",
        "outputId": "575a60e2-b3f5-4b55-9543-96e827415d36",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "data = keras.datasets.mnist\n",
        "(train_images, train_labels), (test_images, test_labels) = data.load_data()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 1s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4_E7k74DBJ4_",
        "colab_type": "text"
      },
      "source": [
        "Wir errinnern uns, dass die Pixelwerte noch nicht in normaliserter Form vorliegen. Wir normalisieren diese also zunächst indem wir durch den maximalen Pixelwert 255 teilen:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6onrRWeHOV8b",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_images = train_images / 255.0\n",
        "test_images = test_images / 255.0"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "teAwiBwgBlgr",
        "colab_type": "text"
      },
      "source": [
        "Zuletzt hatten wir noch das Problem, dass die Labels des Datensatzes einfach nur Zahlen waren.\n",
        "- Das Bild der handgeschriebenen 5 hat das Label `5`.\n",
        "- Wir hätten in diesem Fall jedoch gerne den Vektor `[0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]` als Label, welcher der von unserem Netz gewünschten Ausgabe entspricht. Dieser Vektor hat nur an Stelle 5 (beginnend bei 0) eine 1.\n",
        "\n",
        "Für detailliertere Erklärungen sei an dieser Stelle wieder auf das Video \"Laden und Bearbeiten des MNIST Datensatzes\" verwiesen.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rIKVxQK-C8ii",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "total_classes = 10\n",
        "train_vec_labels = keras.utils.to_categorical(train_labels, total_classes)\n",
        "test_vec_labels = keras.utils.to_categorical(test_labels, total_classes)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vFp_jgGcDRTl",
        "colab_type": "text"
      },
      "source": [
        "#### Entwurf eines Netzes\n",
        "\n",
        "Nun haben wir die Eingabedaten normalisiert und die Labels als Vektoren vorliegen. Wir können also endlich anfangen ein Netz für die Erkennung der handgeschriebenen Zahlen zu bauen! :)  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T-KdN4xbuSAT",
        "colab_type": "text"
      },
      "source": [
        "Wir wollen dafür ein sehr einfaches Netz mit 3 Layern definieren (Eingabelayer, Hidden Layer und Ausgabelayer):\n",
        "- Als **Input-Layer** verwenden wir einen `keras.layers.Flatten` Layer, der die 28x28 Matrizen, die wir als Eingaben erhalten auf $28\\cdot 28 = 784$ Neuronen verteilt\n",
        "- Als nächstes verwenden wir für den **Hidden-Layer** einen `keras.layers.Dense` Layer mit 128 Neuronen, wobei wir 128 als eine gute Anzahl empfinden\n",
        "- Als **Output-Layer** verwenden wir einen `keras.layers.Dense` Layer mit 10 Neuronen, da wir 10 Klassen (Ziffern von 0-9) erkennen wollen\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EEvwJtk7Oppm",
        "colab_type": "code",
        "outputId": "5f58ef20-6881-43bf-a558-86b22d62e439",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "model = keras.Sequential([\n",
        "    keras.layers.Flatten(input_shape=(28, 28)),\n",
        "    keras.layers.Dense(128, activation='sigmoid'), \n",
        "    keras.layers.Dense(10, activation='sigmoid')\n",
        "])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/resource_variable_ops.py:1630: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "If using Keras pass *_constraint arguments to layers.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ii2xQYOQEGZy",
        "colab_type": "text"
      },
      "source": [
        "#### Kompilieren des Netzes\n",
        "\n",
        "Nach dem wir unser Netz definiert haben, müssen wir es *kompilieren*, bevor wir mit dem Training beginnen können.\n",
        "\n",
        "In diesem Schritt legen wir wichtige Parameter für die Trainingsphase fest:\n",
        "- Der **Optimizer** ist der im Training verwendete Lernalgorithmus zur Verbesserung des Netzes. In der letzen Woche haben wir ja bereits *Gradient Descent* und dessen Optimierung *Stochastic Gradient Descent* (SGD, siehe *Deep Dive: Backpropagation*) kennengelernt.\n",
        "- Der **Loss** ist die verwendete Kostenfunktion. Ziel während des Trainings ist es, diese zu minimieren. Wir haben in Woche 1 bereits die Quadratische Fehlerfunktion (*Squared Error*) kennengelernt.\n",
        "- Die **Metrics** sind die während des Trainings ausgewerteten Metrics. Bei allen Klassifikationsproblemen interessiert uns hier die `\"accuracy\"`."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PwbP-ebPhWzu",
        "colab_type": "text"
      },
      "source": [
        "In diesem Beispiel verwenden wir \n",
        "- Den *Stochastic Gradient Descent* (`\"sgd\"`) Lernalgorithmus als unseren Optimizer.\n",
        "- Die `\"mean_squared_error\"` Kostenfunktion, welche im Vergleich zur normalen *Squared Error* Kostenfunktion nicht die Summe, sondern den Mittelwert der Fehler der Ausgabeneuronen berechnet."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mjC8UAqUOyZ6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# sgd = keras.optimizers.SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)\n",
        "model.compile(\n",
        "    optimizer='sgd',\n",
        "    loss='mean_squared_error',\n",
        "    metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5tIC_kvajNM1",
        "colab_type": "text"
      },
      "source": [
        "#### Trainieren des Netzes\n",
        "\n",
        "Jetzt können wir endlich unser Netz tranieren. Dazu verwenden wir die `fit` Methode und übergeben unsere Trainingsbilder als Eingaben mit den dazugehörigen Labels als gewünschte Ausgaben. Die Anzahl der `epochs` gibt an, wie oft das Netz das gesamte Trainingsset zu sehen bekommt. Erhöhen wir die Anzahl der Epochen lassen wir unser Netz länger lernen.  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6QDK4o68O71J",
        "colab_type": "code",
        "outputId": "528e5d3e-ef5f-446e-9be1-34c27f7cc18f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        }
      },
      "source": [
        "model.fit(train_images, train_vec_labels, epochs=10, verbose=True)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 60000 samples\n",
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 5s 89us/sample - loss: 0.0837 - acc: 0.6504\n",
            "Epoch 2/10\n",
            "60000/60000 [==============================] - 4s 62us/sample - loss: 0.0540 - acc: 0.7919\n",
            "Epoch 3/10\n",
            "60000/60000 [==============================] - 4s 62us/sample - loss: 0.0490 - acc: 0.8167\n",
            "Epoch 4/10\n",
            "60000/60000 [==============================] - 4s 62us/sample - loss: 0.0466 - acc: 0.8276\n",
            "Epoch 5/10\n",
            "60000/60000 [==============================] - 4s 62us/sample - loss: 0.0451 - acc: 0.8338\n",
            "Epoch 6/10\n",
            "60000/60000 [==============================] - 4s 61us/sample - loss: 0.0440 - acc: 0.8382\n",
            "Epoch 7/10\n",
            "60000/60000 [==============================] - 4s 60us/sample - loss: 0.0433 - acc: 0.8412\n",
            "Epoch 8/10\n",
            "60000/60000 [==============================] - 4s 59us/sample - loss: 0.0427 - acc: 0.8429\n",
            "Epoch 9/10\n",
            "60000/60000 [==============================] - 4s 59us/sample - loss: 0.0422 - acc: 0.8451\n",
            "Epoch 10/10\n",
            "60000/60000 [==============================] - 4s 59us/sample - loss: 0.0418 - acc: 0.8462\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f5592dbcf28>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MQDgimiQksZe",
        "colab_type": "text"
      },
      "source": [
        "#### Evaluieren des Netzes\n",
        "\n",
        "Bisher hat das Netz nur Trainingsbilder gesehen und damit gelernt. Ziel ist es aber ja, mit unserem Netz neue Bilder von handgeschriebenen Ziffern zu erkennen. Dafür gibt es ja die Testdaten, mit denen wir unser Netz nun auf die Genauigkeit bei ungesehenen Daten überprüfen wollen."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zeujdy_fU0rl",
        "colab_type": "code",
        "outputId": "e7044292-ade0-4ca6-b627-40dde5537042",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "eval_loss, eval_accuracy = model.evaluate(test_images, test_vec_labels, verbose=False)\n",
        "print(\"Model accuracy: %.2f\" % eval_accuracy)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model accuracy: 0.85\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qhJJG4iE7AJG",
        "colab_type": "text"
      },
      "source": [
        "#### Ausgaben des Netzes"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vgb2WafA7HJm",
        "colab_type": "code",
        "outputId": "18caa006-d14b-47f3-f64f-a5c41d8f047b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 699
        }
      },
      "source": [
        "helpers.plot_predictions(model, test_images[:20], labels=test_vec_labels[:20])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdwAAAKqCAYAAACQMO+0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
            "text/plain": [
              "<Figure size 648x864 with 18 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}