ensemble_learning_dl.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Ramyrahmeni/DeepLearning/blob/main/ensemble_learning_dl.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "X1undNA_BUZk"
      },
      "outputs": [],
      "source": [
        "3# generate 2d classification dataset\n",
        "from sklearn.datasets import make_blobs\n",
        "X, y = make_blobs(n_samples=1000, centers=3, n_features=2, cluster_std=2, random_state=2)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0PXgVejFBUZp",
        "outputId": "29c4bf97-4b93-4118-dead-b2d472cdd585"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1000, 2)"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ],
      "source": [
        "X.shape"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y[:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TOcyBzGnCRwt",
        "outputId": "19d0d56e-f25a-471d-fa69-1c310e406cdb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[2],\n",
              "       [2],\n",
              "       [2],\n",
              "       [0],\n",
              "       [1]])"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "iDu4RrJnBUZs",
        "outputId": "b3f814ba-41f5-4f46-9363-0b50e14df5ce"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "from matplotlib import pyplot\n",
        "from pandas import DataFrame\n",
        "# generate 2d classification dataset\n",
        "#X, y = make_blobs(n_samples=1000, centers=3, n_features=2, cluster_std=2, random_state=2)\n",
        "# scatter plot, dots colored by class value\n",
        "df = DataFrame(dict(x=X[:,0], y=X[:,1], label=y))\n",
        "colors = {0:'red', 1:'blue', 2:'green'}\n",
        "fig, ax = pyplot.subplots()\n",
        "grouped = df.groupby('label')\n",
        "for key, group in grouped:\n",
        "    group.plot(ax=ax, kind='scatter', x='x', y='y', label=key, color=colors[key])\n",
        "pyplot.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "nDqCaowPBUZu"
      },
      "outputs": [],
      "source": [
        "from sklearn.preprocessing import OneHotEncoder\n",
        "import numpy as np\n",
        "y=np.reshape(y,(-1,1))\n",
        "\n",
        "enc = OneHotEncoder()\n",
        "y_new=enc.fit_transform(y).toarray()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vYg9MrsSBUZw",
        "outputId": "ad0eaa36-097b-477f-9514-b0e5e91def3b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(1000, 3)\n"
          ]
        }
      ],
      "source": [
        "print(y_new.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QH_z8ALLBUZx",
        "outputId": "0bbbccd5-24a1-4830-90d4-7c55c390e14a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(700, 3) (300, 3)\n"
          ]
        }
      ],
      "source": [
        "import tensorflow.keras.utils as k\n",
        "#y = k.to_categorical(y)\n",
        "# split into train and test\n",
        "n_train = 700\n",
        "trainX, testX = X[:n_train, :], X[n_train:, :]\n",
        "trainy, testy = y_new[:n_train], y_new[n_train:]\n",
        "print(trainy.shape, testy.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OyvzanaKBUZz",
        "outputId": "a10e775a-45c9-4b60-b87d-7ebe8f831589"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(1000, 3)"
            ]
          },
          "execution_count": 28,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y_new.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zxVrv7I3BUZ0",
        "outputId": "26186815-330e-4845-f7b1-0749a45a3f78"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(700, 3) (300, 3)\n"
          ]
        }
      ],
      "source": [
        "import tensorflow.keras.utils as k\n",
        "#y = k.to_categorical(y)\n",
        "# split into train and test\n",
        "\n",
        "n_train = 700\n",
        "trainX, testX = X[:n_train, :], X[n_train:, :]\n",
        "trainy, testy = y_new[:n_train], y_new[n_train:]\n",
        "print(trainy.shape, testy.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bSVf-GiYBUZ1",
        "outputId": "1b37d39c-757d-45b7-bc66-c6a413540a9b"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/layers/core/dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
            "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
          ]
        }
      ],
      "source": [
        "from tensorflow.keras import datasets, layers, models\n",
        "model = models.Sequential()\n",
        "model.add(layers.Dense(25, input_dim=2, activation='relu'))\n",
        "model.add(layers.Dense(3, activation='softmax'))\n",
        "\n",
        "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 196
        },
        "id": "-R_k4pPBBUZ2",
        "outputId": "e524544a-25ea-4887-b579-ace15a2151bb"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1mModel: \"sequential_1\"\u001b[0m\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_1\"</span>\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                        \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape               \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m        Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
              "│ dense_2 (\u001b[38;5;33mDense\u001b[0m)                      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m25\u001b[0m)                  │              \u001b[38;5;34m75\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense_3 (\u001b[38;5;33mDense\u001b[0m)                      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m3\u001b[0m)                   │              \u001b[38;5;34m78\u001b[0m │\n",
              "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                         </span>┃<span style=\"font-weight: bold\"> Output Shape                </span>┃<span style=\"font-weight: bold\">         Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
              "│ dense_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">25</span>)                  │              <span style=\"color: #00af00; text-decoration-color: #00af00\">75</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">3</span>)                   │              <span style=\"color: #00af00; text-decoration-color: #00af00\">78</span> │\n",
              "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m153\u001b[0m (612.00 B)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">153</span> (612.00 B)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m153\u001b[0m (612.00 B)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">153</span> (612.00 B)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        }
      ],
      "source": [
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "u3rrhGOJBUZ3",
        "outputId": "983cd377-c1ad-4d14-8a19-d48175186c88"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(700, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ],
      "source": [
        "trainy.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gzKeYY2rBUZ4"
      },
      "outputs": [],
      "source": [
        "history = model.fit(trainX, trainy, validation_data=(testX, testy), epochs=500, verbose=0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ut6Vt3HSBUZ5",
        "outputId": "b26bdad2-bd9a-4be2-f731-e9f33ec7c122"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train: 0.821, Test: 0.850\n"
          ]
        }
      ],
      "source": [
        "# evaluate the model\n",
        "_, train_acc = model.evaluate(trainX, trainy, verbose=0)\n",
        "_, test_acc = model.evaluate(testX, testy, verbose=0)\n",
        "print('Train: %.3f, Test: %.3f' % (train_acc, test_acc))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xg0Ma6ThBUZ6"
      },
      "outputs": [],
      "source": [
        "from os import makedirs\n",
        "makedirs('models_EL1')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dxtF123wBUZ7",
        "outputId": "5d94b4a1-7980-4e94-e303-66085cfba456"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved models_EL1/model_1.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved models_EL1/model_2.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved models_EL1/model_3.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved models_EL1/model_4.h5\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">Saved models_EL1/model_5.h5\n"
          ]
        }
      ],
      "source": [
        "\n",
        "def fit_model(trainX, trainy):\n",
        "\t# define model\n",
        "\tmodel = models.Sequential()\n",
        "\tmodel.add(layers.Dense(25, input_dim=2, activation='relu'))\n",
        "\tmodel.add(layers.Dense(3, activation='softmax'))\n",
        "\tmodel.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
        "\t# fit model\n",
        "\tmodel.fit(trainX, trainy, epochs=500, verbose=0)\n",
        "\treturn model\n",
        "\n",
        "\n",
        "n_members = 5\n",
        "for i in range(n_members):\n",
        "\t# fit model#initialisation différente à chaque fois\n",
        "\tmodel = fit_model(trainX, trainy)\n",
        "\t# save model\n",
        "\tfilename = 'models_EL1/model_' + str(i + 1) + '.h5'\n",
        "\tmodel.save(filename)\n",
        "\tprint('>Saved %s' % filename)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "x3SQ3Ba-BUZ8"
      },
      "outputs": [],
      "source": [
        "def load_all_models(n_models):\n",
        "\tall_models = list()\n",
        "\tfor i in range(n_models):\n",
        "\t\t# define filename for this ensemble\n",
        "\t\tfilename = 'models_EL1/model_' + str(i + 1) + '.h5'\n",
        "\t\t# load model from file\n",
        "\t\tmodel = models.load_model(filename)\n",
        "\t\t# add to list of members\n",
        "\t\tall_models.append(model)\n",
        "\t\tprint('>loaded %s' % filename)\n",
        "\treturn all_models"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P1Y_pWNmBUZ9",
        "outputId": "97b20ba3-eae4-4f7e-e2cf-d2d415342155"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n",
            "WARNING:absl:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n",
            "WARNING:absl:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n",
            "WARNING:absl:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n",
            "WARNING:absl:Compiled the loaded model, but the compiled metrics have yet to be built. `model.compile_metrics` will be empty until you train or evaluate the model.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            ">loaded models_EL1/model_1.h5\n",
            ">loaded models_EL1/model_2.h5\n",
            ">loaded models_EL1/model_3.h5\n",
            ">loaded models_EL1/model_4.h5\n",
            ">loaded models_EL1/model_5.h5\n",
            "Model Accuracy: 0.847\n",
            "Model Accuracy: 0.853\n",
            "Model Accuracy: 0.840\n",
            "Model Accuracy: 0.843\n",
            "Model Accuracy: 0.850\n"
          ]
        }
      ],
      "source": [
        "all_models=load_all_models(5)\n",
        "for model in all_models:\n",
        "\t#testy_enc = to_categorical(testy)\n",
        "\t_, acc = model.evaluate(testX, testy, verbose=0)\n",
        "\tprint('Model Accuracy: %.3f' % acc)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Dx2IsustBUZ-"
      },
      "outputs": [],
      "source": [
        "def stacked_dataset(members, inputX):\n",
        "\tstackX = None\n",
        "\tfor model in members:\n",
        "\t\t# make prediction\n",
        "\t\tyhat = model.predict(inputX, verbose=0)\n",
        "\t\t# stack predictions into [rows, members, probabilities] 1000,5,3\n",
        "\t\tif stackX is None:\n",
        "\t\t\tstackX = yhat\n",
        "\t\telse:\n",
        "\t\t\tstackX = np.dstack((stackX, yhat))\n",
        "\t# flatten predictions to [rows, members x probabilities] (1000,5*3)\n",
        "\tstackX = stackX.reshape((stackX.shape[0], stackX.shape[1]*stackX.shape[2]))\n",
        "\treturn stackX"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yNiPy2ncBUZ_"
      },
      "outputs": [],
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "def fit_stacked_model(members, inputX, inputy):\n",
        "    # create dataset using ensemble\n",
        "    stackedX = stacked_dataset(members, inputX)\n",
        "    print(stackedX.shape)\n",
        "    # fit standalone model\n",
        "    model = LogisticRegression()\n",
        "    model.fit(stackedX, inputy)\n",
        "    return model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5FvOj84lBUaA",
        "outputId": "d5f9d62f-bf1a-44b2-fb47-0177dd426109"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(700, 2) (300, 2)\n"
          ]
        }
      ],
      "source": [
        "import tensorflow.keras.utils as k\n",
        "#y = k.to_categorical(y)\n",
        "# split into train and test\n",
        "n_train = 700\n",
        "trainX, testX = X[:n_train, :], X[n_train:, :]\n",
        "trainy, testy = y[:n_train], y[n_train:]\n",
        "print(trainX.shape, testX.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ORII-dPeBUaB",
        "outputId": "2572757b-635d-44a8-81f7-aa588c007b27"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1000, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ],
      "source": [
        "y.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pBO7xAQFBUaC",
        "outputId": "2f32aee1-078e-46e2-e043-36c029054152"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(300, 15)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/utils/validation.py:1339: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
            "  y = column_or_1d(y, warn=True)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Stacked Test Accuracy: 0.850\n"
          ]
        }
      ],
      "source": [
        "# make a prediction with the stacked model\n",
        "from sklearn.metrics import accuracy_score\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "def stacked_prediction(members, model, inputX):\n",
        "\t# create dataset using ensemble\n",
        "\tstackedX = stacked_dataset(members, inputX)\n",
        "\t# make a prediction\n",
        "\tyhat = model.predict(stackedX)\n",
        "\treturn yhat\n",
        "\n",
        "\n",
        "model = fit_stacked_model(all_models, testX, testy)\n",
        "# evaluate model on test set\n",
        "yhat = stacked_prediction(all_models, model, testX)\n",
        "acc = accuracy_score(testy, yhat)\n",
        "print('Stacked Test Accuracy: %.3f' % acc)"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "xNXwX6YKH1gj"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.11.5"
    },
    "colab": {
      "provenance": [],
      "include_colab_link": true
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}