Task2_unsupervised_ML.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Task2_unsupervised ML.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyNhbL6huV46h2XuQRIJJmj4",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/shreyash0502/MLmodels/blob/main/Task2_unsupervised_ML.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UsyH1CyJ9Eh2"
      },
      "source": [
        "<h1><b>This is my project for prediction using Unsupervised Machine Learning.</h1>\n",
        "\n",
        "```\n",
        "#Here, I will implement the K-means clustering ML algorithm for my model.\n",
        "```\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "I7b4S0oq-amw"
      },
      "source": [
        "First, we import all the necessary libraries for developing the K-means clustering model.\n",
        "\n",
        "---\n",
        "```\n",
        "# Here, we will use the sci-kit learn python library.\n",
        "# We will also use python libraries- pandas(for data formatting and manipulation) and numpy(for array manipulation).\n",
        "# Finally, I will use matplotlib library for visualization. \n",
        "```\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_msvJGngJJoM"
      },
      "source": [
        "from sklearn.cluster import KMeans\n",
        "from sklearn.model_selection import train_test_split\n",
        "from matplotlib import pyplot as plt\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn import metrics, datasets\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lplxUUJTAlvp"
      },
      "source": [
        "Now, we will import the iris dataset in our project and extract features and labels from it. \n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AmmEElvkgo4i",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 213
        },
        "outputId": "fb51b8e2-3a0f-416d-ee24-818d2e8bce10"
      },
      "source": [
        "iris = datasets.load_iris()\n",
        "iris_df = pd.DataFrame(iris.data, columns = iris.feature_names)\n",
        "display(iris_df.head(5))\n",
        "y = iris.target\n",
        "#print(np.shape(y))\n",
        "data = pd.read_csv(\"http://bit.ly/w-data\")\n",
        "print(\"Data imported successfully!\")\n",
        "X = iris_df.iloc[0:].values"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "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>sepal length (cm)</th>\n",
              "      <th>sepal width (cm)</th>\n",
              "      <th>petal length (cm)</th>\n",
              "      <th>petal width (cm)</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>5.1</td>\n",
              "      <td>3.5</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>4.9</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>4.7</td>\n",
              "      <td>3.2</td>\n",
              "      <td>1.3</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4.6</td>\n",
              "      <td>3.1</td>\n",
              "      <td>1.5</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5.0</td>\n",
              "      <td>3.6</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   sepal length (cm)  sepal width (cm)  petal length (cm)  petal width (cm)\n",
              "0                5.1               3.5                1.4               0.2\n",
              "1                4.9               3.0                1.4               0.2\n",
              "2                4.7               3.2                1.3               0.2\n",
              "3                4.6               3.1                1.5               0.2\n",
              "4                5.0               3.6                1.4               0.2"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Data imported successfully!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2BR9V3oCFA_L"
      },
      "source": [
        "Visualization of data is an important aspect. With visualization, we can understand better and develop human insights like developing trends and patterns in data. \n",
        "\n",
        "Now, we will visualize the data.\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 538
        },
        "id": "f51jEPrOCg4I",
        "outputId": "f6e02fd0-a6d8-4201-e383-f9a749a67443"
      },
      "source": [
        "plt.figure(figsize=(10,8))\n",
        "plt.scatter(X[:, 0], X[:, 1], s = 100, c = 'black')\n",
        "plt.grid()\n",
        "ax = plt.axes()\n",
        "ax.set_facecolor(\"#f3faff\")\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:4: MatplotlibDeprecationWarning: Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
            "  after removing the cwd from sys.path.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e6c996ttBYwt"
      },
      "source": [
        "Choosing the most efficient k, i.e, number of clusters can be tricky and completely depends on the nature of data distribution. So, we will use the \"Elbow method\" to find the correct value of k. Code is shown below: \n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 513
        },
        "id": "OzF8oUlA-B7o",
        "outputId": "1b9397c5-d902-4013-dc6d-bb670c5a8c6f"
      },
      "source": [
        "\n",
        "wcss = []\n",
        "\n",
        "for i in range(1, 11):\n",
        "    kmeans = KMeans(n_clusters = i, init = 'k-means++', \n",
        "                    max_iter = 300, n_init = 10, random_state = 0)\n",
        "    kmeans.fit(X)\n",
        "    wcss.append(kmeans.inertia_)\n",
        "\n",
        "# Plotting the results onto a line graph, \n",
        "# `allowing us to observe 'The elbow'\n",
        "plt.figure(figsize=(10,8))\n",
        "ax = plt.axes()\n",
        "ax.set_facecolor(\"#f3faff\")\n",
        "plt.grid()\n",
        "plt.plot(range(1, 11), wcss)\n",
        "plt.title('The elbow method')\n",
        "plt.xlabel('Number of clusters')\n",
        "plt.ylabel('WCSS') # Within cluster sum of squares\n",
        "plt.show()\n",
        "#print(np.shape(X))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QPcp57xyEq2-"
      },
      "source": [
        "Now, we have successfully imported the data and visualized it. We can now train our K-Means model as shown below. \n",
        "\n",
        "Number of clusters chosen = 3\n",
        "\n",
        "---\n",
        "\n",
        "We will then plot the clusters with different colors. We can't compare the test cases with prediction directly because assigning a number to a cluster is completely arbitrary. So, there are other methods to evaluate our model's performance which will be discussed in a later section."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 483
        },
        "id": "zjLeV3d8h_Lq",
        "outputId": "eff721d3-34b8-4115-f079-36bfb15a54a0"
      },
      "source": [
        "# Applying kmeans to the dataset / Creating the kmeans classifier\n",
        "kmeans = KMeans(n_clusters = 3, init = 'k-means++',\n",
        "                max_iter = 300, n_init = 10, random_state = 0)\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)\n",
        "#kmeans.fit(X_train)\n",
        "y_kmeans = kmeans.fit_predict(X_train)\n",
        "y_pred = kmeans.predict(X_test)\n",
        "plt.figure(figsize=(10,8))\n",
        "ax = plt.axes()\n",
        "ax.set_facecolor(\"#f3faff\")\n",
        "plt.grid()\n",
        "plt.scatter(X_train[y_kmeans == 0, 0], X_train[y_kmeans == 0, 1], \n",
        "            s = 100, c = 'red', label = 'Cluster 1')\n",
        "plt.scatter(X_train[y_kmeans == 1, 0], X_train[y_kmeans == 1, 1], \n",
        "            s = 100, c = 'blue', label = 'Cluster 2')\n",
        "plt.scatter(X_train[y_kmeans == 2, 0], X_train[y_kmeans == 2, 1],\n",
        "            s = 100, c = 'green', label = 'Cluster 3')\n",
        "\n",
        "# Plotting the centroids of the clusters\n",
        "plt.scatter(kmeans.cluster_centers_[:, 0], kmeans.cluster_centers_[:,1], \n",
        "            s = 100, c = 'yellow', label = 'Centroids')\n",
        "plt.legend()\n",
        "plt.show()\n",
        "#print(y_kmeans)\n",
        "#y"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JQNu6vDdDfkb"
      },
      "source": [
        "We can evaluate the performance of the model based on the normalized mutual information score. Code is shown below:\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hf5Nc0tPipIU",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "08a6b786-ab75-4398-af01-38ae73344c03"
      },
      "source": [
        "from sklearn.metrics import normalized_mutual_info_score\n",
        "print(\"accuracy: \", normalized_mutual_info_score(y_test, y_pred))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "accuracy:  0.7762445166809027\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1Dbx_5EhgE5p"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}