document_classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "<a href=\"https://colab.research.google.com/github/porterjenkins/CS180/blob/main/data_science_labs/data_science_lab_11.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a><p><b>After clicking the \"Open in Colab\" link, copy the notebook to your own Google Drive before getting started, or it will not save your work</b></p>"
      ],
      "metadata": {
        "id": "IPir_6bCCFnZ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "# BYU CS 180 Lab 11"
      ],
      "metadata": {
        "id": "_slaQdUGCB0t"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Dependencies for the lab\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import sklearn"
      ],
      "metadata": {
        "id": "Gaema31vjP8f"
      },
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Introduction:\n",
        "You will use the `sklearn` API to train various classifiers. Feel free to use any `sklearn`, `numpy`, or `pandas` function to help you with the lab.\n",
        "\n",
        "The 20 newsgroups dataset comprises around 18000 newsgroups posts on 20 topics. The 20 categories are:\n",
        "```\n",
        "['alt.atheism',\n",
        " 'comp.graphics',\n",
        " 'comp.os.ms-windows.misc',\n",
        " 'comp.sys.ibm.pc.hardware',\n",
        " 'comp.sys.mac.hardware',\n",
        " 'comp.windows.x',\n",
        " 'misc.forsale',\n",
        " 'rec.autos',\n",
        " 'rec.motorcycles',\n",
        " 'rec.sport.baseball',\n",
        " 'rec.sport.hockey',\n",
        " 'sci.crypt',\n",
        " 'sci.electronics',\n",
        " 'sci.med',\n",
        " 'sci.space',\n",
        " 'soc.religion.christian',\n",
        " 'talk.politics.guns',\n",
        " 'talk.politics.mideast',\n",
        " 'talk.politics.misc',\n",
        " 'talk.religion.misc']\n",
        "```\n",
        "The posts are raw text. For example:\n",
        "\n",
        "**Post:**\n",
        "```\n",
        "From: lerxst@wam.umd.edu (where's my thing)\n",
        "Subject: WHAT car is this!?\n",
        "Nntp-Posting-Host: rac3.wam.umd.edu\n",
        "Organization: University of Maryland, College Park\n",
        "Lines: 15\n",
        "\n",
        " I was wondering if anyone out there could enlighten me on this car I saw\n",
        "the other day. It was a 2-door sports car, looked to be from the late 60s/\n",
        "early 70s. It was called a Bricklin. The doors were really small. In addition,\n",
        "the front bumper was separate from the rest of the body. This is \n",
        "all I know. If anyone can tellme a model name, engine specs, years\n",
        "of production, where this car is made, history, or whatever info you\n",
        "have on this funky looking car, please e-mail.\n",
        "\n",
        "Thanks,\n",
        "- IL\n",
        "   ---- brought to you by your neighborhood Lerxst ----\n",
        "```\n",
        "\n",
        "**Target:**\n",
        "```\n",
        "'rec.autos'\n",
        "```\n",
        "\n",
        "Your task will be to build a classifier to predict the category of posts. You will do the following in this notebook:"
      ],
      "metadata": {
        "id": "ct7fnkcnCL8O"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Exercise 1: Retrieve the Data\n",
        "Download `fetch_20newsgroups` dataset"
      ],
      "metadata": {
        "id": "PUat5xRAcdrC"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.datasets import fetch_20newsgroups\n",
        "data = fetch_20newsgroups()\n",
        "docs = data['data']\n",
        "targets = data[\"target\"]\n",
        "names = data.target_names"
      ],
      "metadata": {
        "id": "j1m2KIHShNdC"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Print off the first document and its target classification to make sure its been downloaded correctly "
      ],
      "metadata": {
        "id": "BXDLQKs3hUM8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(docs[0])\n",
        "print(names[targets[0]])"
      ],
      "metadata": {
        "id": "-hfFbcRohgMf",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d996a81d-c302-4f68-a040-767bf145d765"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "From: lerxst@wam.umd.edu (where's my thing)\n",
            "Subject: WHAT car is this!?\n",
            "Nntp-Posting-Host: rac3.wam.umd.edu\n",
            "Organization: University of Maryland, College Park\n",
            "Lines: 15\n",
            "\n",
            " I was wondering if anyone out there could enlighten me on this car I saw\n",
            "the other day. It was a 2-door sports car, looked to be from the late 60s/\n",
            "early 70s. It was called a Bricklin. The doors were really small. In addition,\n",
            "the front bumper was separate from the rest of the body. This is \n",
            "all I know. If anyone can tellme a model name, engine specs, years\n",
            "of production, where this car is made, history, or whatever info you\n",
            "have on this funky looking car, please e-mail.\n",
            "\n",
            "Thanks,\n",
            "- IL\n",
            "   ---- brought to you by your neighborhood Lerxst ----\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "rec.autos\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Exercise 2: Understand the Data"
      ],
      "metadata": {
        "id": "RSUun4WVcl17"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Perform some exploratory analysis on your data. Specifically, plot a bar chart showing distribution of the categories (classes)"
      ],
      "metadata": {
        "id": "xJZbee3ihnqV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import seaborn as sns\n",
        "import numpy as np\n",
        "\n",
        "counts = np.bincount(targets)\n",
        "ax = sns.barplot(x = names, y = counts)\n",
        "ax.set_xticklabels(ax.get_xticklabels(), rotation=90)\n"
      ],
      "metadata": {
        "id": "veGuUMbEhj5-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 960
        },
        "outputId": "f970f507-5e94-4b67-e1f1-f4ea6e0a8b3c"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[Text(0, 0, 'alt.atheism'),\n",
              " Text(1, 0, 'comp.graphics'),\n",
              " Text(2, 0, 'comp.os.ms-windows.misc'),\n",
              " Text(3, 0, 'comp.sys.ibm.pc.hardware'),\n",
              " Text(4, 0, 'comp.sys.mac.hardware'),\n",
              " Text(5, 0, 'comp.windows.x'),\n",
              " Text(6, 0, 'misc.forsale'),\n",
              " Text(7, 0, 'rec.autos'),\n",
              " Text(8, 0, 'rec.motorcycles'),\n",
              " Text(9, 0, 'rec.sport.baseball'),\n",
              " Text(10, 0, 'rec.sport.hockey'),\n",
              " Text(11, 0, 'sci.crypt'),\n",
              " Text(12, 0, 'sci.electronics'),\n",
              " Text(13, 0, 'sci.med'),\n",
              " Text(14, 0, 'sci.space'),\n",
              " Text(15, 0, 'soc.religion.christian'),\n",
              " Text(16, 0, 'talk.politics.guns'),\n",
              " Text(17, 0, 'talk.politics.mideast'),\n",
              " Text(18, 0, 'talk.politics.misc'),\n",
              " Text(19, 0, 'talk.religion.misc')]"
            ]
          },
          "metadata": {},
          "execution_count": 5
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Exercise 3: Engineer the Data\n",
        "Extract features from raw text. You are free to be creative here. There is no correct answer. Try and think of the features (input) that are most predictive of the category. Some helpful sklearn feature engineering code can be found here (I'd recommend reading through the documentation before diving in and trying to get these tools to work):\n",
        "\n",
        "* [CountVectorizer](https://scikit-learn.org/stable/modules/generated/sklearn.feature_extraction.text.CountVectorizer.html#sklearn.feature_extraction.text.CountVectorizer)\n",
        "* [TFIDF Text Features](https://scikit-learn.org/stable/modules/generated/sklearn.feature_extraction.text.TfidfVectorizer.html)"
      ],
      "metadata": {
        "id": "CBmTXeqpcq5W"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Enter code here to engineer the data\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "vectorizer = CountVectorizer()\n",
        "X = vectorizer.fit(docs)\n",
        "vectorizer.get_feature_names_out()"
      ],
      "metadata": {
        "id": "M18gziAZifj0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8de4dcd4-119c-4a00-a785-eca9e110df93"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['00', '000', '0000', ..., 'íålittin', 'ñaustin', 'ýé'],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X = X.toarray()"
      ],
      "metadata": {
        "id": "TGi7MxKBVGCi"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vectorizer2 = TfidfVectorizer()\n",
        "X2 = vectorizer2.fit(docs)\n",
        "vectorizer2.get_feature_names_out()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ym1cHUyGOoBl",
        "outputId": "b08f6efd-f334-4e11-da41-bb872059fbe0"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['00', '000', '0000', ..., 'íålittin', 'ñaustin', 'ýé'],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X2 = X2.toarray()"
      ],
      "metadata": {
        "id": "pIZ0fESMWNzM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Exercise 4: Split the Data\n",
        "\n",
        "Split your data into train and test splits. Use an 80%/20% split rule for train and test respectively. I'd recommend using the function from the last lab"
      ],
      "metadata": {
        "id": "JN4r2NSpc5hQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#Enter code here to split the dataset\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(docs, targets, test_size= 0.2, random_state = 42)\n"
      ],
      "metadata": {
        "id": "C6uiqrStkxww"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Exercise 5: Train the Model"
      ],
      "metadata": {
        "id": "eW6QTIUxcz_G"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Using sklearn, train the following three classifiers using the training set:\n",
        "* KNN\n",
        "* Decision Tree\n",
        "* Logistic Regression"
      ],
      "metadata": {
        "id": "mNY2czLak_pA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model using the KNN Cluster Algorithm\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "train_feat = vectorizer.fit_transform(X_train)\n",
        "test_feat = vectorizer.transform(X_test)\n",
        "knn = KNeighborsClassifier(n_neighbors = 5)\n",
        "knn.fit(X=train_feat, y = y_train)\n",
        "y_hat = knn.predict(X=test_feat)"
      ],
      "metadata": {
        "id": "MIaBp38BlEzu"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model using the Decision Tree Algorithm\n",
        "from sklearn import tree\n",
        "classifier = tree.DecisionTreeClassifier(\n",
        "    criterion='entropy',\n",
        "    max_depth=10,\n",
        "    splitter='best'\n",
        ")\n",
        "classifier.fit(X=train_feat, y=y_train)\n",
        "y_hat_tree = classifier.predict(test_feat)"
      ],
      "metadata": {
        "id": "xSINn2VklIH9"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model using the Logistic Regression \n",
        "from sklearn.linear_model import LogisticRegression\n",
        "model = LogisticRegression(max_iter = 35)\n",
        "model.fit(train_feat, y_train)\n",
        "y_hat_log = model.predict(test_feat)"
      ],
      "metadata": {
        "id": "ZayZjHO1lNXG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "84f6f4d6-f509-4eb8-8e6f-a44f32e0c337"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/sklearn/linear_model/_logistic.py:458: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
            "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
            "\n",
            "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
            "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
            "Please also refer to the documentation for alternative solver options:\n",
            "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
            "  n_iter_i = _check_optimize_result(\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Exercise 6: Evaluate the Model\n",
        "Compare the performance of your classifiers on the test set. Discuss which one performs the best."
      ],
      "metadata": {
        "id": "gRC3UrQjdOtM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the KNN Cluster model\n",
        "from sklearn.metrics import accuracy_score\n",
        "accuracy_score(y_test, y_hat)"
      ],
      "metadata": {
        "id": "fviUlnNZl_Ek",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7b7dca8f-9c57-4302-aa95-be00f4ce6902"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.482545293857711"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the Decision Tree model\n",
        "accuracy_score(y_test, y_hat_tree)"
      ],
      "metadata": {
        "id": "NFysD0Krl9-n",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fee44956-2d69-4fa6-e00c-82f421b263a6"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.312417145382236"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the Logistic Regression model\n",
        "accuracy_score(y_test, y_hat_log)"
      ],
      "metadata": {
        "id": "FVjIpzzxl8E-",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "dfe8e390-31f2-4f54-b0b0-3a7b7541fcd9"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8055678303137428"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Discuss which one performs the best:"
      ],
      "metadata": {
        "id": "ofWWlon9mqUw"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "The logistic regression model has the highest accuracy score, so it performed the best when fitting the test set. "
      ],
      "metadata": {
        "id": "rLE2qUqpmu0P"
      }
    }
  ]
}