random_forest_classification.ipynb · Tony Han/unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF

1月18日，北京，聊聊2025如何加入技术开发？

加入 Gitee

与超过 1200万开发者一起发现、参与优秀开源项目，私有仓库也完全免费：）

文件

该仓库未声明开源许可证文件（LICENSE），使用请关注具体项目描述及其代码上游依赖。

克隆/下载

random_forest_classification.ipynb 148.62 KB

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "random_forest_classification.ipynb",
      "provenance": [],
      "mount_file_id": "1oiiCt5NsqSSt2vR2-r7Yz0OnmsS3_Y-z",
      "authorship_tag": "ABX9TyPxayV7e8anBws/YV9CidKj",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/monsund/unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF/blob/master/random_forest_classification.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iPVeUdv-flGv"
      },
      "source": [
        "import pandas as pd"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YNsYWWGhftCp"
      },
      "source": [
        "input_file = '/content/drive/My Drive/NLP/with_swaraj/Data/data_for_statistical_classification_1.csv'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C6_2lFOUf8sO"
      },
      "source": [
        "df = pd.read_csv(input_file)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dIBM9V_dgFrf",
        "outputId": "7cfc8a69-a2a0-4289-9c88-36972768fd0c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Unnamed: 0</th>\n",
              "      <th>url</th>\n",
              "      <th>username</th>\n",
              "      <th>title</th>\n",
              "      <th>about</th>\n",
              "      <th>dominant_topic</th>\n",
              "      <th>topic_prob</th>\n",
              "      <th>keywords</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>https://www.linkedin.com/in/lokesh-kumar-xess-...</td>\n",
              "      <td>lokesh-kumar-xess-54814068</td>\n",
              "      <td>Marketing &amp; Strategy at Mobistreak</td>\n",
              "      <td>I am a Google Adwords and Google Analytics cer...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.184839</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>77</td>\n",
              "      <td>https://www.linkedin.com/in/jdellisanti</td>\n",
              "      <td>jdellisanti</td>\n",
              "      <td>Product Innovation, Strategy and Marketing Exe...</td>\n",
              "      <td>Innovation, Strategy and Marketing Leader at t...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.175725</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>78</td>\n",
              "      <td>https://www.linkedin.com/in/niharmanwatkar</td>\n",
              "      <td>niharmanwatkar</td>\n",
              "      <td>Founder  The Banana Design Company | Helping b...</td>\n",
              "      <td>Specialties: Digital Marketing, Marketing &amp; Co...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.207002</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>99</td>\n",
              "      <td>https://www.linkedin.com/in/gregarious</td>\n",
              "      <td>gregarious</td>\n",
              "      <td>Serial Entrepreneur, Startup Mentor, Innovatio...</td>\n",
              "      <td>Gregarious “Greg” Narain has been at the foref...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.161111</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>143</td>\n",
              "      <td>https://www.linkedin.com/in/nadyakhoja</td>\n",
              "      <td>nadyakhoja</td>\n",
              "      <td>Chief Growth Officer at Venngage</td>\n",
              "      <td>Experienced Head Of Marketing with a demonstra...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.161481</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Unnamed: 0  ...                                           keywords\n",
              "0           0  ...  ['marketing', 'digital', 'brand', 'medium', 'c...\n",
              "1          77  ...  ['marketing', 'digital', 'brand', 'medium', 'c...\n",
              "2          78  ...  ['marketing', 'digital', 'brand', 'medium', 'c...\n",
              "3          99  ...  ['marketing', 'digital', 'brand', 'medium', 'c...\n",
              "4         143  ...  ['marketing', 'digital', 'brand', 'medium', 'c...\n",
              "\n",
              "[5 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YcfSrNNaHiGI",
        "outputId": "9ce85f7a-84a6-46a3-e940-2261a473c290",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "print(f'unique topic numbers: {df.dominant_topic.unique()}','\\n')\n",
        "topic_dic = {5: 'marketing domain', 8:'related to product visionnaire', 7:'related to data domain', 10:'software domain',14: 'entrepreneur or working with entrepreneur', \n",
        "             11:'related to investment; startups looking for investment', 3:'managerial field', 12:'business,strategic growth,revenue', 4:'hr or people management'}\n",
        "topic_dic"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "unique topic numbers: [ 5  8  7 10 14 11  3 12  4] \n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{3: 'managerial field',\n",
              " 4: 'hr or people management',\n",
              " 5: 'marketing domain',\n",
              " 7: 'related to data domain',\n",
              " 8: 'related to product visionnaire',\n",
              " 10: 'software domain',\n",
              " 11: 'related to investment; startups looking for investment',\n",
              " 12: 'business,strategic growth,revenue',\n",
              " 14: 'entrepreneur or working with entrepreneur'}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w-mezAedYt-u"
      },
      "source": [
        "# **Remove Punctuation**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PslC-zpLYQsq"
      },
      "source": [
        "# define punctuation\n",
        "punctuations = '\\'!()-[]{};:\\\"\\,<>./?#$%^&*_~'"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1B9BqylmaLmg"
      },
      "source": [
        "#Function for removing punctuation and lowering case\n",
        "def remove_punct(text):\n",
        "  for char in text:\n",
        "    if char in punctuations:\n",
        "      text = text.lower().replace(char,' ')    #lowering case\n",
        "      text = text.replace('  ', ' ')\n",
        "  return text"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eBikD64SanB3",
        "outputId": "ccf1d514-67b3-4e99-8b3c-c88869df6b49",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 210
        }
      },
      "source": [
        "df['about_cleaned']= df.about.apply(lambda x: remove_punct(x))\n",
        "df.head(3)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Unnamed: 0</th>\n",
              "      <th>url</th>\n",
              "      <th>username</th>\n",
              "      <th>title</th>\n",
              "      <th>about</th>\n",
              "      <th>dominant_topic</th>\n",
              "      <th>topic_prob</th>\n",
              "      <th>keywords</th>\n",
              "      <th>about_cleaned</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>https://www.linkedin.com/in/lokesh-kumar-xess-...</td>\n",
              "      <td>lokesh-kumar-xess-54814068</td>\n",
              "      <td>Marketing &amp; Strategy at Mobistreak</td>\n",
              "      <td>I am a Google Adwords and Google Analytics cer...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.184839</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>i am a google adwords and google analytics cer...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>77</td>\n",
              "      <td>https://www.linkedin.com/in/jdellisanti</td>\n",
              "      <td>jdellisanti</td>\n",
              "      <td>Product Innovation, Strategy and Marketing Exe...</td>\n",
              "      <td>Innovation, Strategy and Marketing Leader at t...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.175725</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>innovation strategy and marketing leader at th...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>78</td>\n",
              "      <td>https://www.linkedin.com/in/niharmanwatkar</td>\n",
              "      <td>niharmanwatkar</td>\n",
              "      <td>Founder  The Banana Design Company | Helping b...</td>\n",
              "      <td>Specialties: Digital Marketing, Marketing &amp; Co...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.207002</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>specialties digital marketing marketing commun...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Unnamed: 0  ...                                      about_cleaned\n",
              "0           0  ...  i am a google adwords and google analytics cer...\n",
              "1          77  ...  innovation strategy and marketing leader at th...\n",
              "2          78  ...  specialties digital marketing marketing commun...\n",
              "\n",
              "[3 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "828lcEGxpBAJ"
      },
      "source": [
        "# **Remove Stopwords**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BY3d-ytMowtT",
        "outputId": "167292eb-1e9e-48e9-856f-8b71ea6d26de",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        }
      },
      "source": [
        "import nltk\n",
        "nltk.download('punkt')\n",
        "from nltk.corpus import stopwords\n",
        "nltk.download('stopwords')\n",
        "from nltk.tokenize import word_tokenize"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Unzipping tokenizers/punkt.zip.\n",
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yd0xuAYTpzSJ"
      },
      "source": [
        "stop_words = set(stopwords.words('english'))\n",
        "additional_stop_words = set(['technology', 'development', 'experience', 'project', 'company'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NCRfTHrX7Txx"
      },
      "source": [
        "stop_words = stop_words.union(additional_stop_words)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lMGSaDrdpO95"
      },
      "source": [
        "# Function removing stopwords-----------\n",
        "def remove_stopwords(text):\n",
        "  word_tokens = word_tokenize(text)\n",
        "  filtered_text = [w for w in word_tokens if not w in stop_words]\n",
        "  data = (' '.join(filtered_text))\n",
        "  return data"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ze5URi9hqOAi"
      },
      "source": [
        "df['about_cleaned'] = df.about_cleaned.apply(lambda x: remove_stopwords(x))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GOaBGUgZgeRZ",
        "outputId": "72db55a0-0a43-4981-e1f2-6155d95aff72",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Unnamed: 0</th>\n",
              "      <th>url</th>\n",
              "      <th>username</th>\n",
              "      <th>title</th>\n",
              "      <th>about</th>\n",
              "      <th>dominant_topic</th>\n",
              "      <th>topic_prob</th>\n",
              "      <th>keywords</th>\n",
              "      <th>about_cleaned</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>https://www.linkedin.com/in/lokesh-kumar-xess-...</td>\n",
              "      <td>lokesh-kumar-xess-54814068</td>\n",
              "      <td>Marketing &amp; Strategy at Mobistreak</td>\n",
              "      <td>I am a Google Adwords and Google Analytics cer...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.184839</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>google adwords google analytics certified ppc ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>77</td>\n",
              "      <td>https://www.linkedin.com/in/jdellisanti</td>\n",
              "      <td>jdellisanti</td>\n",
              "      <td>Product Innovation, Strategy and Marketing Exe...</td>\n",
              "      <td>Innovation, Strategy and Marketing Leader at t...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.175725</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>innovation strategy marketing leader nexus med...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>78</td>\n",
              "      <td>https://www.linkedin.com/in/niharmanwatkar</td>\n",
              "      <td>niharmanwatkar</td>\n",
              "      <td>Founder  The Banana Design Company | Helping b...</td>\n",
              "      <td>Specialties: Digital Marketing, Marketing &amp; Co...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.207002</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>specialties digital marketing marketing commun...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>99</td>\n",
              "      <td>https://www.linkedin.com/in/gregarious</td>\n",
              "      <td>gregarious</td>\n",
              "      <td>Serial Entrepreneur, Startup Mentor, Innovatio...</td>\n",
              "      <td>Gregarious “Greg” Narain has been at the foref...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.161111</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>gregarious “ greg ” narain forefront social me...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>143</td>\n",
              "      <td>https://www.linkedin.com/in/nadyakhoja</td>\n",
              "      <td>nadyakhoja</td>\n",
              "      <td>Chief Growth Officer at Venngage</td>\n",
              "      <td>Experienced Head Of Marketing with a demonstra...</td>\n",
              "      <td>5</td>\n",
              "      <td>0.161481</td>\n",
              "      <td>['marketing', 'digital', 'brand', 'medium', 'c...</td>\n",
              "      <td>experienced head marketing demonstrated histor...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Unnamed: 0  ...                                      about_cleaned\n",
              "0           0  ...  google adwords google analytics certified ppc ...\n",
              "1          77  ...  innovation strategy marketing leader nexus med...\n",
              "2          78  ...  specialties digital marketing marketing commun...\n",
              "3          99  ...  gregarious “ greg ” narain forefront social me...\n",
              "4         143  ...  experienced head marketing demonstrated histor...\n",
              "\n",
              "[5 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TqNMF-Bq5yQ4"
      },
      "source": [
        "# **Train Test Split**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-hQjlZJ953Aw"
      },
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "X = df.drop(['dominant_topic'],axis=1)\n",
        "y = df['dominant_topic']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SBYJIbXD6EiG"
      },
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(X,y, test_size=0.2, random_state=42)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Pjmq9f1LiYzR"
      },
      "source": [
        "# **TDIDF Vectorization**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YtzUJts2iOho"
      },
      "source": [
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "tfidf_vec = TfidfVectorizer()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MazPMRT6i3hI"
      },
      "source": [
        "X_traincv = tfidf_vec.fit_transform(X_train.about.to_list())\n",
        "X_testcv = tfidf_vec.transform(X_test.about.to_list())"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x_jsM7I3-jak"
      },
      "source": [
        "X_traincv_array = X_traincv.toarray()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oU2o2bFBhGlf"
      },
      "source": [
        "# **Save TFIDF Model**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rDGvF30-hPhQ"
      },
      "source": [
        "import pickle\n",
        "with open(\"/content/drive/My Drive/NLP/with_swaraj/model/tfidf_model_profile_classification.pkl\", \"wb\") as fp:   #Pickling\n",
        "  pickle.dump(tfidf_vec, fp)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "skAFD_Q7TP-O"
      },
      "source": [
        "# **Clustering**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oNCQeQ8BTS0U"
      },
      "source": [
        "# from sklearn.cluster import KMeans"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1SdVlRDuTTqo"
      },
      "source": [
        "# kmeans_model = KMeans(n_clusters=9, init='k-means++', max_iter=100) \n",
        "# # X = kmeans_model.fit(model_pvdm.docvecs.doctag_syn0)\n",
        "# x = kmeans_model.fit(X_traincv_array)\n",
        "# labels = kmeans_model.labels_.tolist()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SfTAS2QPTtKY"
      },
      "source": [
        "# kmeans_prediction = kmeans_model.fit_predict(X_traincv_array)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mx2R8Pgx-69y"
      },
      "source": [
        "# **Scatter plot**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TZA142Py-9qN"
      },
      "source": [
        "# # from sklearn.decomposition import PCA\n",
        "\n",
        "# pca = PCA(n_components=2).fit(X_traincv_array)\n",
        "# datapoint = pca.transform(X_traincv_array)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oA8mHozzAvsk",
        "outputId": "a3ff0fb6-4449-4aa3-94c1-48a32b747608",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 320
        }
      },
      "source": [
        "# import matplotlib.pyplot as plt\n",
        "# %matplotlib inline\n",
        "# plt.figure(figsize=(10, 5))\n",
        "# # label1 = ['#FFFF00', '#008000', '#0000FF', '#800080', '#800450', '#800FF0', '#112328', '#620000', '#FF00FF', 'orange','#00FF66', '#302B54', '#CDAD00', '#D0FAEE', '#D41A1F']\n",
        "# label1 =['red', 'green', 'yellow', 'blue','black','pink','brown','orange','violet']\n",
        "# color = [label1[i] for i in labels]\n",
        "# plt.scatter(datapoint[:, 0], datapoint[:, 1], c=color)\n",
        "# centroids = kmeans_model.cluster_centers_\n",
        "# centroidpoint = pca.transform(centroids)\n",
        "# plt.scatter(centroidpoint[:, 0], centroidpoint[:, 1], marker='^', s=150, c='#000000')\n",
        "# plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "td-np0rHBBmK"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W6_95iBbBqLJ"
      },
      "source": [
        "# **Random Forest**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FPhv9EvRDDEx",
        "outputId": "c21e2d46-3f6d-43f5-bee3-5e1b89241161",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "#  importing the Random Forest model\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "model = RandomForestClassifier(n_estimators=100)\n",
        "model.fit(X_traincv,y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,\n",
              "                       criterion='gini', max_depth=None, max_features='auto',\n",
              "                       max_leaf_nodes=None, max_samples=None,\n",
              "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                       min_samples_leaf=1, min_samples_split=2,\n",
              "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
              "                       n_jobs=None, oob_score=False, random_state=None,\n",
              "                       verbose=0, warm_start=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "huoTVM93EK2l"
      },
      "source": [
        "# Predicted results\n",
        "y_result = model.predict(X_testcv)\n",
        "y_prob = model.predict_proba(X_testcv)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Yx0dmw1y9Y2B",
        "outputId": "5d05e1ea-9ce6-4875-fa63-67996fe8385a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_prob[1]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.02, 0.  , 0.04, 0.05, 0.02, 0.01, 0.69, 0.05, 0.12])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d4EiMcB2EWdY",
        "outputId": "bcd3f979-bb9b-4029-a7f3-8a2661854fe2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "# Accuracy check\n",
        "from sklearn.metrics import accuracy_score, recall_score, precision_score\n",
        "\n",
        "accuracy = accuracy_score(y_result,y_test)\n",
        "recall = recall_score(y_test, y_result, average='micro')\n",
        "precision = precision_score(y_test, y_result, average='weighted')\n",
        "\n",
        "print(f'accuracy: {accuracy}' ,'\\n' )\n",
        "print(f'recall: {recall}' ,'\\n' )\n",
        "print(f'precision: {precision}' ,'\\n' )"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "accuracy: 0.8644688644688645 \n",
            "\n",
            "recall: 0.8644688644688645 \n",
            "\n",
            "precision: 0.8744262704442309 \n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xNf8Okg49mfM",
        "outputId": "24722b12-21ef-4049-dce0-fce91d15bef3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "topic_dic"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{3: 'managerial field',\n",
              " 4: 'hr or people management',\n",
              " 5: 'marketing domain',\n",
              " 7: 'related to data domain',\n",
              " 8: 'related to product visionnaire',\n",
              " 10: 'software domain',\n",
              " 11: 'related to investment; startups looking for investment',\n",
              " 12: 'business,strategic growth,revenue',\n",
              " 14: 'entrepreneur or working with entrepreneur'}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DOxNXRscRu7u"
      },
      "source": [
        "# **Function for finding Probability of Each Class**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F5Nm50VcMuRV"
      },
      "source": [
        "def get_class_prob(y_prob_list):\n",
        "  topic_prob = []\n",
        "  for i in range(len(y_prob_list)):\n",
        "    topic3_prob = y_prob_list[i][0]\n",
        "    topic4_prob = y_prob_list[i][1]\n",
        "    topic5_prob = y_prob_list[i][2]\n",
        "    topic7_prob = y_prob_list[i][3]\n",
        "    topic8_prob = y_prob_list[i][4]\n",
        "    topic10_prob = y_prob_list[i][5]\n",
        "    topic11_prob = y_prob_list[i][6]\n",
        "    topic12_prob = y_prob_list[i][7]\n",
        "    topic14_prob = y_prob_list[i][8]\n",
        "    \n",
        "    dic = {'managerial field':topic3_prob, 'hr or people management':topic4_prob, 'marketing domain':topic5_prob, 'data science domain':topic7_prob,\n",
        "          'product visionnaire, management related field':topic8_prob, 'software domain':topic10_prob, 'related to investment,startups looking for investment':topic11_prob,\n",
        "          'business,strategic growth,revenue':topic12_prob, 'entrepreneur or working with entrepreneur':topic14_prob}\n",
        "    sort_dic = sorted(dic.items(), key=lambda x: x[1], reverse=True)\n",
        "    topic_prob.append(sort_dic)\n",
        "  return topic_prob  "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PPnE8S0wPbtb"
      },
      "source": [
        "topic_prob = get_class_prob(y_prob[:len(y_prob)])\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s5hRpTV48MWb",
        "outputId": "0688e872-12ff-416f-dccf-1fd2fc53d004",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "topic_prob[1]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[('related to investment,startups looking for investment', 0.69),\n",
              " ('entrepreneur or working with entrepreneur', 0.12),\n",
              " ('data science domain', 0.05),\n",
              " ('business,strategic growth,revenue', 0.05),\n",
              " ('marketing domain', 0.04),\n",
              " ('managerial field', 0.02),\n",
              " ('product visionnaire, management related field', 0.02),\n",
              " ('software domain', 0.01),\n",
              " ('hr or people management', 0.0)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 36
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GqMss1GgVO-h"
      },
      "source": [
        "# **Export Result**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4kAR_BwiWl22"
      },
      "source": [
        "df_final = X_test"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "drvxlV3BWsmO",
        "outputId": "6b6240a4-f0f3-42f6-9824-723c4e52da5c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 323
        }
      },
      "source": [
        "df_final['actual'] = y_test\n",
        "df_final['prediction'] = y_result\n",
        "df_final['class_prob'] = topic_prob"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  \"\"\"Entry point for launching an IPython kernel.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  \n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:3: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  This is separate from the ipykernel package so we can avoid doing imports until\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Toy8NZPtVR8P"
      },
      "source": [
        "df_final.to_csv('/content/drive/My Drive/NLP/with_swaraj/Data/random_forest_classification_result_1.csv', index=False)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "69QVglIQTT3q"
      },
      "source": [
        "# **Confusion Matrix**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QzJmmS2DEX7I",
        "outputId": "b8898f9d-eb04-4f5e-cf70-7b06e2c1b4ad",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "from sklearn.metrics import confusion_matrix\n",
        "confusion_matrix(y_test,y_result)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[27,  0,  2,  0,  1,  1,  0,  0,  0],\n",
              "       [ 1, 15,  1,  0,  0,  0,  0,  0,  0],\n",
              "       [ 0,  0, 43,  0,  0,  0,  1,  0,  0],\n",
              "       [ 2,  0,  0, 28,  0,  5,  0,  0,  0],\n",
              "       [ 0,  0,  0,  0, 30,  2,  0,  0,  0],\n",
              "       [ 0,  0,  0,  0,  0, 44,  0,  0,  0],\n",
              "       [ 0,  0,  0,  0,  0,  0, 28,  0,  0],\n",
              "       [ 1,  0,  5,  2,  3,  0,  0, 12,  2],\n",
              "       [ 0,  0,  3,  1,  1,  0,  3,  0,  9]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 32
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zKm2MzavY4Sg"
      },
      "source": [
        "# **Save Model**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pZj2Wm-3ZFCz"
      },
      "source": [
        "import pickle\n",
        "\n",
        "with open(\"/content/drive/My Drive/NLP/with_swaraj/model/ranforest_model.pkl\", \"wb\") as fp:   #Pickling\n",
        "  pickle.dump(model, fp)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5G-xzANZaKbO"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}

一键复制编辑原始数据按行查看历史

提交于 2020-10-28 20:22 . random_forest_classification

马建仓 AI 助手

尝试更多

代码解读

代码找茬

代码优化

https://gitee.com/tony-han/unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF.git

[email protected]:tony-han/unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF.git

tony-han

unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF

master

Tony Han/unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF

简介

发行版

贡献者

近期动态

Tony Han/unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF .gitee-modal { width: 500px !important; }

简介

发行版

贡献者

近期动态

搜索帮助

Tony Han/unsupervised-classification-of-Linkedin-Profiles-using-KMeans-LDA-TFIDF