{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "AFHtqXrAOsg0",
        "outputId": "a38eb973-c892-4ce1-aa81-91a63640a8f4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "5/5 [==============================] - 1s 69ms/step - loss: 1.0193 - acc: 0.4222 - val_loss: 0.4962 - val_acc: 1.0000\n",
            "Epoch 2/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.9108 - acc: 0.5556 - val_loss: 0.8713 - val_acc: 1.0000\n",
            "Epoch 3/50\n",
            "5/5 [==============================] - 0s 16ms/step - loss: 0.8500 - acc: 0.6296 - val_loss: 0.5394 - val_acc: 1.0000\n",
            "Epoch 4/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.7936 - acc: 0.6296 - val_loss: 0.6457 - val_acc: 1.0000\n",
            "Epoch 5/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.7479 - acc: 0.6296 - val_loss: 0.6321 - val_acc: 1.0000\n",
            "Epoch 6/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.7151 - acc: 0.6296 - val_loss: 0.6199 - val_acc: 1.0000\n",
            "Epoch 7/50\n",
            "5/5 [==============================] - 0s 13ms/step - loss: 0.6862 - acc: 0.6296 - val_loss: 0.6487 - val_acc: 1.0000\n",
            "Epoch 8/50\n",
            "5/5 [==============================] - 0s 16ms/step - loss: 0.6610 - acc: 0.6963 - val_loss: 0.6296 - val_acc: 1.0000\n",
            "Epoch 9/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.6368 - acc: 0.7259 - val_loss: 0.5884 - val_acc: 1.0000\n",
            "Epoch 10/50\n",
            "5/5 [==============================] - 0s 18ms/step - loss: 0.6128 - acc: 0.7333 - val_loss: 0.6335 - val_acc: 1.0000\n",
            "Epoch 11/50\n",
            "5/5 [==============================] - 0s 15ms/step - loss: 0.5930 - acc: 0.8593 - val_loss: 0.7045 - val_acc: 0.9333\n",
            "Epoch 12/50\n",
            "5/5 [==============================] - 0s 15ms/step - loss: 0.5754 - acc: 0.9481 - val_loss: 0.7206 - val_acc: 0.9333\n",
            "Epoch 13/50\n",
            "5/5 [==============================] - 0s 19ms/step - loss: 0.5563 - acc: 0.9481 - val_loss: 0.6533 - val_acc: 1.0000\n",
            "Epoch 14/50\n",
            "5/5 [==============================] - 0s 19ms/step - loss: 0.5363 - acc: 0.9556 - val_loss: 0.6609 - val_acc: 0.9333\n",
            "Epoch 15/50\n",
            "5/5 [==============================] - 0s 16ms/step - loss: 0.5245 - acc: 0.9556 - val_loss: 0.6424 - val_acc: 0.9333\n",
            "Epoch 16/50\n",
            "5/5 [==============================] - 0s 14ms/step - loss: 0.5082 - acc: 0.9556 - val_loss: 0.7203 - val_acc: 0.6000\n",
            "Epoch 17/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.4987 - acc: 0.9185 - val_loss: 0.5814 - val_acc: 1.0000\n",
            "Epoch 18/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.4844 - acc: 0.9778 - val_loss: 0.5839 - val_acc: 1.0000\n",
            "Epoch 19/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.4740 - acc: 0.9778 - val_loss: 0.5918 - val_acc: 1.0000\n",
            "Epoch 20/50\n",
            "5/5 [==============================] - 0s 13ms/step - loss: 0.4661 - acc: 0.9481 - val_loss: 0.5877 - val_acc: 1.0000\n",
            "Epoch 21/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.4559 - acc: 0.9481 - val_loss: 0.4701 - val_acc: 1.0000\n",
            "Epoch 22/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.4494 - acc: 0.9481 - val_loss: 0.4756 - val_acc: 1.0000\n",
            "Epoch 23/50\n",
            "5/5 [==============================] - 0s 19ms/step - loss: 0.4403 - acc: 0.9630 - val_loss: 0.5246 - val_acc: 1.0000\n",
            "Epoch 24/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.4329 - acc: 0.9407 - val_loss: 0.6121 - val_acc: 0.9333\n",
            "Epoch 25/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.4224 - acc: 0.9556 - val_loss: 0.5385 - val_acc: 1.0000\n",
            "Epoch 26/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.4168 - acc: 0.9778 - val_loss: 0.5839 - val_acc: 0.9333\n",
            "Epoch 27/50\n",
            "5/5 [==============================] - 0s 19ms/step - loss: 0.4078 - acc: 0.9481 - val_loss: 0.4825 - val_acc: 1.0000\n",
            "Epoch 28/50\n",
            "5/5 [==============================] - 0s 16ms/step - loss: 0.4070 - acc: 0.9556 - val_loss: 0.6198 - val_acc: 0.9333\n",
            "Epoch 29/50\n",
            "5/5 [==============================] - 0s 17ms/step - loss: 0.3950 - acc: 0.9481 - val_loss: 0.5656 - val_acc: 0.9333\n",
            "Epoch 30/50\n",
            "5/5 [==============================] - 0s 18ms/step - loss: 0.3934 - acc: 0.9704 - val_loss: 0.6974 - val_acc: 0.5333\n",
            "Epoch 31/50\n",
            "5/5 [==============================] - 0s 16ms/step - loss: 0.3853 - acc: 0.9333 - val_loss: 0.4559 - val_acc: 1.0000\n",
            "Epoch 32/50\n",
            "5/5 [==============================] - 0s 11ms/step - loss: 0.3821 - acc: 0.9630 - val_loss: 0.4834 - val_acc: 1.0000\n",
            "Epoch 33/50\n",
            "5/5 [==============================] - 0s 10ms/step - loss: 0.3784 - acc: 0.9481 - val_loss: 0.3687 - val_acc: 1.0000\n",
            "Epoch 34/50\n",
            "5/5 [==============================] - 0s 15ms/step - loss: 0.3785 - acc: 0.9185 - val_loss: 0.5095 - val_acc: 1.0000\n",
            "Epoch 35/50\n",
            "5/5 [==============================] - 0s 14ms/step - loss: 0.3653 - acc: 0.9630 - val_loss: 0.5972 - val_acc: 0.9333\n",
            "Epoch 36/50\n",
            "5/5 [==============================] - 0s 11ms/step - loss: 0.3592 - acc: 0.9407 - val_loss: 0.4897 - val_acc: 1.0000\n",
            "Epoch 37/50\n",
            "5/5 [==============================] - 0s 11ms/step - loss: 0.3546 - acc: 0.9556 - val_loss: 0.8035 - val_acc: 0.2667\n",
            "Epoch 38/50\n",
            "5/5 [==============================] - 0s 15ms/step - loss: 0.3621 - acc: 0.9259 - val_loss: 0.6529 - val_acc: 0.6000\n",
            "Epoch 39/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.3476 - acc: 0.9407 - val_loss: 0.4267 - val_acc: 1.0000\n",
            "Epoch 40/50\n",
            "5/5 [==============================] - 0s 11ms/step - loss: 0.3425 - acc: 0.9778 - val_loss: 0.6261 - val_acc: 0.7333\n",
            "Epoch 41/50\n",
            "5/5 [==============================] - 0s 15ms/step - loss: 0.3463 - acc: 0.9407 - val_loss: 0.7577 - val_acc: 0.3333\n",
            "Epoch 42/50\n",
            "5/5 [==============================] - 0s 13ms/step - loss: 0.3392 - acc: 0.9333 - val_loss: 0.5413 - val_acc: 0.9333\n",
            "Epoch 43/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.3357 - acc: 0.9333 - val_loss: 0.3666 - val_acc: 1.0000\n",
            "Epoch 44/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.3351 - acc: 0.9556 - val_loss: 0.6111 - val_acc: 0.7333\n",
            "Epoch 45/50\n",
            "5/5 [==============================] - 0s 10ms/step - loss: 0.3232 - acc: 0.9259 - val_loss: 0.3965 - val_acc: 1.0000\n",
            "Epoch 46/50\n",
            "5/5 [==============================] - 0s 11ms/step - loss: 0.3196 - acc: 0.9778 - val_loss: 0.4509 - val_acc: 1.0000\n",
            "Epoch 47/50\n",
            "5/5 [==============================] - 0s 16ms/step - loss: 0.3154 - acc: 0.9778 - val_loss: 0.7080 - val_acc: 0.4667\n",
            "Epoch 48/50\n",
            "5/5 [==============================] - 0s 15ms/step - loss: 0.3146 - acc: 0.9259 - val_loss: 0.6699 - val_acc: 0.5333\n",
            "Epoch 49/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.3107 - acc: 0.9333 - val_loss: 0.5593 - val_acc: 0.8667\n",
            "Epoch 50/50\n",
            "5/5 [==============================] - 0s 12ms/step - loss: 0.3035 - acc: 0.9481 - val_loss: 0.6260 - val_acc: 0.7333\n",
            "5/5 [==============================] - 0s 3ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f7788f5a830>"
            ]
          },
          "metadata": {},
          "execution_count": 33
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "from tensorflow.keras import Sequential\n",
        "from tensorflow.keras.layers import Dense\n",
        "\n",
        "from sklearn.datasets import load_iris\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "data = load_iris()\n",
        "x_train, y_train = data.data, data.target\n",
        "\n",
        "model = Sequential()\n",
        "model.add(Dense(32, activation='relu', input_dim = 4))\n",
        "model.add(Dense(8, activation='relu'))\n",
        "model.add(Dense(3, activation='softmax'))\n",
        "\n",
        "model.compile(loss=\"sparse_categorical_crossentropy\", optimizer=\"SGD\", metrics=['acc'])\n",
        "\n",
        "history = model.fit(x_train, y_train, epochs = 50, validation_split = 0.1)\n",
        "\n",
        "y_hat = model.predict(x_train)\n",
        "classes = y_hat.argmax(axis=1)\n",
        "acc = sum(y_train == classes)/len(x_train)\n",
        "\n",
        "os_x = np.arange(50)+1\n",
        "loss = history.history['loss']\n",
        "vloss = history.history['val_loss']\n",
        "acc_list = history.history['acc']\n",
        "vacc_list = history.history['val_acc']\n",
        "\n",
        "fig, ax = plt.subplots(2,1)\n",
        "ax[0].plot(os_x, loss, label=\"train\")\n",
        "ax[0].plot(os_x, vloss, label=\"train\")\n",
        "ax[0].legend()\n",
        "ax[1].plot(os_x,acc_list, label=\"train\")\n",
        "ax[1].plot(os_x,vacc_list,label=\"test\")\n",
        "ax[1].legend(loc='lower right')\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    }
  ]
}