supervised_2.ipynb

{
  "cells": [
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "view-in-github"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/CYaiche/Machine_Learning/blob/master/projet5/embedding_topic_modeling.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "Exy2TPsxHFl5"
      },
      "source": [
        "\n",
        "# Supervised topic modeling : NN approach\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "In this notebook, I tested two different embeddings (third one, BERT in the third notebook) which are **Word2Vec** and **USE** embedding. \n",
        "I will train Word2Vec on my corpus and use a pre-trained USE embedding. \n",
        "\n",
        "For the classification, I trained a multi-layer perceptron. \n",
        "I evaluate both model with the same score as the first notebook : the average precision score and the jaccard score. I also plot confusion matrix for some specific classes. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 74,
      "metadata": {
        "id": "Nebb9sgrHSLV"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "The autoreload extension is already loaded. To reload it, use:\n",
            "  %reload_ext autoreload\n"
          ]
        }
      ],
      "source": [
        "%load_ext autoreload\n",
        "%autoreload 2\n",
        "from  tm_common import *\n",
        "IN_COLAB, output_dir = tm_get_working_config()\n",
        "import os, sys\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import gensim.corpora as corpora\n",
        "from gensim.models import KeyedVectors\n",
        "from gensim.models import Word2Vec\n",
        "\n",
        "from sklearn.preprocessing import MultiLabelBinarizer\n",
        "from keras.utils import pad_sequences\n",
        "from sklearn.metrics import jaccard_score, average_precision_score\n",
        "from sklearn.metrics import classification_report\n",
        "import tensorflow as tf\n",
        "import tensorflow_hub as hub\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import (\n",
        "    Flatten,\n",
        "    Embedding,\n",
        "    Dense,\n",
        "    Input,\n",
        "    Embedding,\n",
        "    GlobalAveragePooling1D,\n",
        ")\n",
        "\n",
        "from sentence_transformers import SentenceTransformer"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Python version used :  3.8.10 \n",
            "Tensorflow  :  2.10.1\n"
          ]
        }
      ],
      "source": [
        "print(\"Python version used : \", sys.version[:7])\n",
        "print(\"Tensorflow  : \", tf.version.VERSION)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 75,
      "metadata": {
        "id": "vawUwgJuG6E0"
      },
      "outputs": [],
      "source": [
        "X_train, X_test, y_train, y_test = tm_load_train_test_set(output_dir)\n",
        "\n",
        "X_corpus_train, X_corpus_test = tm_load_train_test_set(output_dir, option=\"raw_corpus\")\n",
        "label_list = tm_get_label_list(output_dir)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 76,
      "metadata": {},
      "outputs": [],
      "source": [
        "X_train, y_train = tm_get_subset(X_train, y_train)\n",
        "X_test, y_test = tm_get_subset(X_test, y_test)\n",
        "X_corpus_train = tm_get_subsetX(X_corpus_train)\n",
        "X_corpus_test = tm_get_subsetX(X_corpus_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 77,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "1545\n",
            "1545\n",
            "6182\n",
            "6182\n",
            "6182\n",
            "1545\n"
          ]
        }
      ],
      "source": [
        "print( len(X_test) ) \n",
        "print( len(y_test) )\n",
        "print( len(y_train))\n",
        "print( len(X_train))\n",
        "print( len(X_corpus_train))\n",
        "print( len(X_corpus_test))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 55,
      "metadata": {},
      "outputs": [],
      "source": [
        "X_corpus_train, X_corpus_test = tm_get_subset(X_corpus_train, X_corpus_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "J-xWVEn28vC7"
      },
      "outputs": [],
      "source": [
        "y_train_b, y_test_b = tm_multilabel_binarizer(y_train, y_test)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "uS-D3tkeHOPZ"
      },
      "source": [
        "# Neural Networks\n",
        "\n",
        "Multilayers perceptron (MLP) for multi-label classification\n"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "BJxW23-VJqpn"
      },
      "source": [
        "# Embedding preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "if85zWvcMoHq"
      },
      "outputs": [],
      "source": [
        "X_train_list = [ x.tolist() for x in X_train]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GF6stGYZMmbe",
        "outputId": "d7064a51-5c35-4df6-9ef1-f9a17d6e3743"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Build & train Word2Vec model ...\n"
          ]
        }
      ],
      "source": [
        "#  Word2Vec model creation\n",
        "w2v_size=300\n",
        "w2v_window=5\n",
        "w2v_min_count=1\n",
        "w2v_epochs=100\n",
        "maxlen = 24 # adapt to length of sentences\n",
        "\n",
        "sentences = X_train_list\n",
        "\n",
        "print(\"Build & train Word2Vec model ...\")\n",
        "w2v_model = Word2Vec(min_count=w2v_min_count, window=w2v_window,\n",
        "                                                vector_size=w2v_size,\n",
        "                                                seed=42,\n",
        "                                                workers=1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "KYe3tvkgMGS7"
      },
      "outputs": [],
      "source": [
        "try:\n",
        "    model_vectors = KeyedVectors.load(f\"{output_dir}model_vector.wv\", mmap=\"r\")\n",
        "except:\n",
        "    w2v_model.build_vocab(sentences)\n",
        "    w2v_model.train(sentences, total_examples=w2v_model.corpus_count, epochs=w2v_epochs)\n",
        "    model_vectors = w2v_model.wv\n",
        "\n",
        "    model_vectors.save(f\"{output_dir}model_vector.wv\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Vocabulary size: 370103\n",
            "Word2Vec trained\n"
          ]
        }
      ],
      "source": [
        "w2v_words = model_vectors.index_to_key\n",
        "print(\"Vocabulary size: %i\" % len(w2v_words))\n",
        "print(\"Word2Vec trained\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DMBiPMxDQsjW",
        "outputId": "865db21a-6ed0-4fb3-81d2-f8f1652abc97"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Word2Vec<vocab=0, vector_size=300, alpha=0.025>\n"
          ]
        }
      ],
      "source": [
        "print(w2v_model)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "t-xufgfQallL"
      },
      "outputs": [],
      "source": [
        "d = dict(zip(label_list, range(0,len(label_list))))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "jaLWWIQLOJBS"
      },
      "outputs": [],
      "source": [
        "id2word = corpora.Dictionary(X_train)\n",
        "x_train_ids = [ id2word.doc2idx(tokens) for tokens in X_train]\n",
        "\n",
        "x_test_ids = [ id2word.doc2idx(tokens) for tokens in X_test  ]\n",
        "# Filter out -1 from x_test_ids\n",
        "filtered_x_test_ids = [[word_id for word_id in sentence if word_id != -1] for sentence in x_test_ids]\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-NCXNAT_RrMA",
        "outputId": "6c994365-07fe-4629-cfd5-91ad74ebc196"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "max_length : 1862\n"
          ]
        }
      ],
      "source": [
        "max_length = np.max([ len(x) for x in x_train_ids])\n",
        "print(f\"max_length : {max_length}\")"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "ETr4guPjpjau"
      },
      "source": [
        "# Embedding matrix"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CZM-9CW-qch7",
        "outputId": "9235cfae-1754-4aec-eeac-0380f83d3416"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "300"
            ]
          },
          "execution_count": 14,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "len(model_vectors[0])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "087wFGULpmS9",
        "outputId": "5656f95f-8d51-456d-e982-8ab9e702e207"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Embedding matrix: (370103, 300)\n"
          ]
        }
      ],
      "source": [
        "word2vec_vocab_size = len(model_vectors.key_to_index)\n",
        "w2vec_dim = len(model_vectors[0])\n",
        "word2vec_embedding_matrix = np.zeros((word2vec_vocab_size, w2vec_dim))\n",
        "\n",
        "for word in model_vectors.key_to_index:\n",
        "    embedding_vector = model_vectors[word]\n",
        "    if embedding_vector is not None:\n",
        "        idx = model_vectors.key_to_index[word]\n",
        "        word2vec_embedding_matrix[idx] = embedding_vector\n",
        "\n",
        "print(\"Embedding matrix: %s\" % str(word2vec_embedding_matrix.shape))"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "nfo8-bOfbMLW"
      },
      "source": [
        "# Apply padding"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "r91-eP4gY51k"
      },
      "outputs": [],
      "source": [
        "x_train_pad = pad_sequences(x_train_ids, maxlen=max_length, padding='pre',value=0)\n",
        "x_test_pad = pad_sequences(filtered_x_test_ids, maxlen=max_length, padding='pre',value=0)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fbkX2vkYQWml",
        "outputId": "3937b14a-5ee8-48ad-959f-f435e1b6d576"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " embedding (Embedding)       (None, 1862, 300)         111030900 \n",
            "                                                                 \n",
            " global_average_pooling1d (G  (None, 300)              0         \n",
            " lobalAveragePooling1D)                                          \n",
            "                                                                 \n",
            " dense (Dense)               (None, 30)                9030      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 111,039,930\n",
            "Trainable params: 111,039,930\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "None\n"
          ]
        }
      ],
      "source": [
        "model = Sequential()\n",
        "# model.add(Input(shape=(1,max_length),dtype='float64'))\n",
        "model.add(Embedding( word2vec_vocab_size, w2vec_dim,weights=[word2vec_embedding_matrix],  input_length=max_length))\n",
        "model.add(GlobalAveragePooling1D())\n",
        "model.add(Dense(30,activation='sigmoid'))\n",
        "print(model.summary())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "GEpi_Spe3KSF"
      },
      "outputs": [],
      "source": [
        "model.compile(optimizer='adam', loss='binary_crossentropy',metrics=['accuracy'])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qp97pRuR8bu-",
        "outputId": "9942339f-43c2-41ba-92d0-b863a750ae20"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "49/49 [==============================] - 56s 1s/step - loss: 0.2914 - accuracy: 0.1186\n",
            "Epoch 2/5\n",
            "49/49 [==============================] - 55s 1s/step - loss: 0.1996 - accuracy: 0.1600\n",
            "Epoch 3/5\n",
            "49/49 [==============================] - 55s 1s/step - loss: 0.1984 - accuracy: 0.1576\n",
            "Epoch 4/5\n",
            "49/49 [==============================] - 55s 1s/step - loss: 0.1987 - accuracy: 0.1567\n",
            "Epoch 5/5\n",
            "49/49 [==============================] - 56s 1s/step - loss: 0.1985 - accuracy: 0.1529\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x1b8d870ae20>"
            ]
          },
          "execution_count": 19,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "model.fit(x_train_pad, y_train_b, batch_size=128, epochs=5)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "194/194 [==============================] - 4s 20ms/step\n"
          ]
        }
      ],
      "source": [
        "y_pred = model.predict(x_train_pad)\n",
        "threshold_test = tm_test_threshold(y_train_b, y_pred)  \n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "tm_plot_threshold_test(threshold_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "0.09999999999999346"
            ]
          },
          "execution_count": 22,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "\n",
        "max_config = threshold_test.query('precision == precision.max()')\n",
        "t = max_config[\"threshold\"]\n",
        "best_t =  t.values[0]\n",
        "\n",
        "best_t"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "O8ofX6_a__Ne",
        "outputId": "1bc97b10-fccb-414e-bea5-dec4ea5a2e04"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "49/49 [==============================] - 1s 19ms/step\n"
          ]
        }
      ],
      "source": [
        "y_pred = model.predict(x_test_pad)\n",
        "y_pred_nn = (y_pred > best_t).astype(np.float32)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "ezkkFTFV9e8Q"
      },
      "outputs": [],
      "source": [
        "\n",
        "precision_w2v      = average_precision_score(y_test_b, y_pred_nn, average='micro')\n",
        "jaccard_score_w2v = jaccard_score(y_test_b, y_pred_nn, average='micro')"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "jl4XyvqgK4h0"
      },
      "source": [
        "The vocab is the number of unique words in my train data.\n",
        "The size is the dimension in output of my embedding."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 45,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 3000x1000 with 3 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plot_confusion_matrix(y_test_b, y_pred_nn, label_list)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                precision    recall  f1-score   support\n",
            "\n",
            "            c#       0.16      1.00      0.28       251\n",
            "          java       0.15      1.00      0.27       236\n",
            "           c++       0.11      1.00      0.20       176\n",
            "    javascript       0.13      1.00      0.23       205\n",
            "        python       0.00      0.00      0.00       174\n",
            "       android       0.16      0.35      0.22       134\n",
            "           ios       0.05      0.01      0.02       110\n",
            "          .net       0.07      0.01      0.01       123\n",
            "           php       0.00      0.00      0.00        79\n",
            "             c       0.06      0.01      0.02        86\n",
            "          html       0.00      0.00      0.00        98\n",
            "        jquery       0.00      0.00      0.00        66\n",
            "   performance       0.25      0.01      0.02        77\n",
            "           css       0.00      0.00      0.00        68\n",
            "   objective-c       0.00      0.00      0.00        39\n",
            "         c++11       0.00      0.00      0.00        46\n",
            "       asp.net       0.00      0.00      0.00        35\n",
            "       node.js       0.00      0.00      0.00        51\n",
            "           sql       0.00      0.00      0.00        45\n",
            "     algorithm       0.00      0.00      0.00        37\n",
            "         linux       0.00      0.00      0.00        42\n",
            "        spring       0.00      0.00      0.00        41\n",
            "multithreading       0.00      0.00      0.00        38\n",
            "        iphone       0.00      0.00      0.00        30\n",
            "         swift       0.00      0.00      0.00        35\n",
            "         mysql       0.00      0.00      0.00        45\n",
            "       windows       1.00      0.03      0.06        31\n",
            "          json       0.00      0.00      0.00        40\n",
            "        arrays       0.00      0.00      0.00        31\n",
            "         xcode       0.00      0.00      0.00        26\n",
            "\n",
            "     micro avg       0.14      0.37      0.20      2495\n",
            "     macro avg       0.07      0.15      0.04      2495\n",
            "  weighted avg       0.09      0.37      0.10      2495\n",
            "   samples avg       0.14      0.40      0.20      2495\n",
            "\n"
          ]
        }
      ],
      "source": [
        "print(classification_report(y_test_b, y_pred_nn, target_names=label_list, digits=2))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "The support is clearly correlated to the precision here, which is less than the SVM one, the embedding clearly lacks of training. For next embedding, I will use a pre-train one. "
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# USE : Universal Sentence Encoder \n",
        "encodes into high-dimensional vectors, trained on variety of data and output 512 dimensional vector"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# apply padding to words "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 36,
      "metadata": {},
      "outputs": [],
      "source": [
        "embed = hub.load(\"https://tfhub.dev/google/universal-sentence-encoder/4\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 78,
      "metadata": {},
      "outputs": [],
      "source": [
        "X_train_use_embedding = embed(X_corpus_train.tolist())\n",
        "X_test_use_embedding = embed(X_corpus_test.tolist())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 79,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_7 (Dense)             (None, 256)               131328    \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 128)               32896     \n",
            "                                                                 \n",
            " dense_9 (Dense)             (None, 30)                3870      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 168,094\n",
            "Trainable params: 168,094\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "use_model = Sequential()\n",
        "use_model.add(Dense(256, activation='relu', input_dim=512))\n",
        "use_model.add(Dense(128, activation='relu'))\n",
        "use_model.add(Dense(30,activation='sigmoid'))\n",
        "# Compile the model\n",
        "use_model.compile(optimizer='adam', loss='binary_crossentropy',metrics=['accuracy'])\n",
        "use_model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 80,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "49/49 [==============================] - 1s 3ms/step - loss: 0.3518 - accuracy: 0.1301\n",
            "Epoch 2/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.1786 - accuracy: 0.4397\n",
            "Epoch 3/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.1419 - accuracy: 0.6250\n",
            "Epoch 4/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.1111 - accuracy: 0.6978\n",
            "Epoch 5/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0960 - accuracy: 0.7273\n",
            "Epoch 6/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0878 - accuracy: 0.7329\n",
            "Epoch 7/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0823 - accuracy: 0.7407\n",
            "Epoch 8/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0784 - accuracy: 0.7383\n",
            "Epoch 9/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0755 - accuracy: 0.7422\n",
            "Epoch 10/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0732 - accuracy: 0.7407\n",
            "Epoch 11/100\n",
            "49/49 [==============================] - 0s 4ms/step - loss: 0.0713 - accuracy: 0.7409\n",
            "Epoch 12/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0697 - accuracy: 0.7412\n",
            "Epoch 13/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0679 - accuracy: 0.7347\n",
            "Epoch 14/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0666 - accuracy: 0.7413\n",
            "Epoch 15/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0652 - accuracy: 0.7409\n",
            "Epoch 16/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0641 - accuracy: 0.7420\n",
            "Epoch 17/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0631 - accuracy: 0.7412\n",
            "Epoch 18/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0620 - accuracy: 0.7413\n",
            "Epoch 19/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0607 - accuracy: 0.7426\n",
            "Epoch 20/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0597 - accuracy: 0.7467\n",
            "Epoch 21/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0586 - accuracy: 0.7468\n",
            "Epoch 22/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0578 - accuracy: 0.7473\n",
            "Epoch 23/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0566 - accuracy: 0.7509\n",
            "Epoch 24/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0555 - accuracy: 0.7512\n",
            "Epoch 25/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0547 - accuracy: 0.7522\n",
            "Epoch 26/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0535 - accuracy: 0.7583\n",
            "Epoch 27/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0526 - accuracy: 0.7577\n",
            "Epoch 28/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0517 - accuracy: 0.7608\n",
            "Epoch 29/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0506 - accuracy: 0.7640\n",
            "Epoch 30/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0498 - accuracy: 0.7596\n",
            "Epoch 31/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0487 - accuracy: 0.7671\n",
            "Epoch 32/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0479 - accuracy: 0.7659\n",
            "Epoch 33/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0467 - accuracy: 0.7701\n",
            "Epoch 34/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0457 - accuracy: 0.7697\n",
            "Epoch 35/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0447 - accuracy: 0.7729\n",
            "Epoch 36/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0437 - accuracy: 0.7747\n",
            "Epoch 37/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0429 - accuracy: 0.7794\n",
            "Epoch 38/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0421 - accuracy: 0.7811\n",
            "Epoch 39/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0409 - accuracy: 0.7858\n",
            "Epoch 40/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0399 - accuracy: 0.7818\n",
            "Epoch 41/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0388 - accuracy: 0.7896\n",
            "Epoch 42/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0378 - accuracy: 0.7897\n",
            "Epoch 43/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0370 - accuracy: 0.7933\n",
            "Epoch 44/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0360 - accuracy: 0.7952\n",
            "Epoch 45/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0353 - accuracy: 0.7970\n",
            "Epoch 46/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0341 - accuracy: 0.7965\n",
            "Epoch 47/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0334 - accuracy: 0.7986\n",
            "Epoch 48/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0323 - accuracy: 0.8038\n",
            "Epoch 49/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0316 - accuracy: 0.8062\n",
            "Epoch 50/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0306 - accuracy: 0.8091\n",
            "Epoch 51/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0298 - accuracy: 0.8082\n",
            "Epoch 52/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0290 - accuracy: 0.8130\n",
            "Epoch 53/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0279 - accuracy: 0.8073\n",
            "Epoch 54/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0273 - accuracy: 0.8117\n",
            "Epoch 55/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0263 - accuracy: 0.8156\n",
            "Epoch 56/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0257 - accuracy: 0.8172\n",
            "Epoch 57/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0247 - accuracy: 0.8166\n",
            "Epoch 58/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0238 - accuracy: 0.8203\n",
            "Epoch 59/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0231 - accuracy: 0.8217\n",
            "Epoch 60/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0222 - accuracy: 0.8172\n",
            "Epoch 61/100\n",
            "49/49 [==============================] - 0s 4ms/step - loss: 0.0215 - accuracy: 0.8232\n",
            "Epoch 62/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0209 - accuracy: 0.8235\n",
            "Epoch 63/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0199 - accuracy: 0.8230\n",
            "Epoch 64/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0193 - accuracy: 0.8250\n",
            "Epoch 65/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0184 - accuracy: 0.8221\n",
            "Epoch 66/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0177 - accuracy: 0.8269\n",
            "Epoch 67/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0171 - accuracy: 0.8227\n",
            "Epoch 68/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0165 - accuracy: 0.8276\n",
            "Epoch 69/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0158 - accuracy: 0.8269\n",
            "Epoch 70/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0152 - accuracy: 0.8303\n",
            "Epoch 71/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0147 - accuracy: 0.8280\n",
            "Epoch 72/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0139 - accuracy: 0.8293\n",
            "Epoch 73/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0133 - accuracy: 0.8313\n",
            "Epoch 74/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0127 - accuracy: 0.8326\n",
            "Epoch 75/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0121 - accuracy: 0.8327\n",
            "Epoch 76/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0117 - accuracy: 0.8310\n",
            "Epoch 77/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0111 - accuracy: 0.8277\n",
            "Epoch 78/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0107 - accuracy: 0.8306\n",
            "Epoch 79/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0102 - accuracy: 0.8334\n",
            "Epoch 80/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0096 - accuracy: 0.8335\n",
            "Epoch 81/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0091 - accuracy: 0.8353\n",
            "Epoch 82/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0086 - accuracy: 0.8321\n",
            "Epoch 83/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0083 - accuracy: 0.8339\n",
            "Epoch 84/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0078 - accuracy: 0.8344\n",
            "Epoch 85/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0075 - accuracy: 0.8334\n",
            "Epoch 86/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0070 - accuracy: 0.8339\n",
            "Epoch 87/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0067 - accuracy: 0.8339\n",
            "Epoch 88/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0064 - accuracy: 0.8348\n",
            "Epoch 89/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0061 - accuracy: 0.8303\n",
            "Epoch 90/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0058 - accuracy: 0.8348\n",
            "Epoch 91/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0054 - accuracy: 0.8348\n",
            "Epoch 92/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0052 - accuracy: 0.8347\n",
            "Epoch 93/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0050 - accuracy: 0.8365\n",
            "Epoch 94/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0047 - accuracy: 0.8379\n",
            "Epoch 95/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0044 - accuracy: 0.8357\n",
            "Epoch 96/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0041 - accuracy: 0.8368\n",
            "Epoch 97/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0040 - accuracy: 0.8344\n",
            "Epoch 98/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0038 - accuracy: 0.8337\n",
            "Epoch 99/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0036 - accuracy: 0.8348\n",
            "Epoch 100/100\n",
            "49/49 [==============================] - 0s 3ms/step - loss: 0.0034 - accuracy: 0.8339\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x1b8da4ddee0>"
            ]
          },
          "execution_count": 80,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "use_model.fit(X_train_use_embedding, y_train_b, batch_size = 128, epochs=100)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 81,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "4/4 [==============================] - 0s 2ms/step\n"
          ]
        }
      ],
      "source": [
        "y_pred = use_model.predict(X_train_use_embedding[:100])\n",
        "threshold_test_use = tm_test_threshold(y_train_b[:100], y_pred)  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 82,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "tm_plot_threshold_test(threshold_test_use)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 83,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "0.3981071705534692"
            ]
          },
          "execution_count": 83,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "max_config = threshold_test_use.query('precision == precision.max()')\n",
        "t = max_config[\"threshold\"]\n",
        "best_t =  t.values[0]\n",
        "best_t"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 84,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(1545, 30)"
            ]
          },
          "execution_count": 84,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y_test_b.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 85,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "TensorShape([1545, 512])"
            ]
          },
          "execution_count": 85,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "X_test_use_embedding.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 86,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "49/49 [==============================] - 0s 941us/step\n"
          ]
        }
      ],
      "source": [
        "y_pred_use = use_model.predict(X_test_use_embedding)\n",
        "y_pred_use = (y_pred_use > best_t).astype(np.float32)\n",
        "precision_use      = average_precision_score(y_test_b, y_pred_use, average='micro')\n",
        "jaccard_score_use = jaccard_score(y_test_b, y_pred_use, average='micro')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 89,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 3000x1000 with 3 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plot_confusion_matrix(y_test_b, y_pred_use, label_list)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 90,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                precision    recall  f1-score   support\n",
            "\n",
            "            c#       0.75      0.74      0.75       251\n",
            "          java       0.76      0.81      0.78       236\n",
            "           c++       0.75      0.75      0.75       176\n",
            "    javascript       0.77      0.75      0.76       205\n",
            "        python       0.90      0.86      0.88       174\n",
            "       android       0.82      0.83      0.82       134\n",
            "           ios       0.79      0.74      0.76       110\n",
            "          .net       0.47      0.38      0.42       123\n",
            "           php       0.71      0.71      0.71        79\n",
            "             c       0.62      0.60      0.61        86\n",
            "          html       0.64      0.54      0.59        98\n",
            "        jquery       0.62      0.71      0.66        66\n",
            "   performance       0.49      0.36      0.42        77\n",
            "           css       0.78      0.66      0.71        68\n",
            "   objective-c       0.42      0.46      0.44        39\n",
            "         c++11       0.42      0.22      0.29        46\n",
            "       asp.net       0.31      0.23      0.26        35\n",
            "       node.js       0.53      0.65      0.58        51\n",
            "           sql       0.63      0.71      0.67        45\n",
            "     algorithm       0.57      0.54      0.56        37\n",
            "         linux       0.55      0.38      0.45        42\n",
            "        spring       0.73      0.73      0.73        41\n",
            "multithreading       0.57      0.53      0.55        38\n",
            "        iphone       0.25      0.20      0.22        30\n",
            "         swift       0.50      0.26      0.34        35\n",
            "         mysql       0.64      0.56      0.60        45\n",
            "       windows       0.38      0.39      0.38        31\n",
            "          json       0.69      0.50      0.58        40\n",
            "        arrays       0.50      0.35      0.42        31\n",
            "         xcode       0.71      0.38      0.50        26\n",
            "\n",
            "     micro avg       0.69      0.65      0.67      2495\n",
            "     macro avg       0.61      0.55      0.57      2495\n",
            "  weighted avg       0.68      0.65      0.66      2495\n",
            "   samples avg       0.68      0.69      0.65      2495\n",
            "\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "c:\\dev\\topic_modelling\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1344: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in samples with no predicted labels. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ],
      "source": [
        "print(classification_report(y_test_b, y_pred_use, target_names=label_list, digits=2))"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# Evaluation "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 91,
      "metadata": {},
      "outputs": [],
      "source": [
        "avg_precision_scores = [precision_w2v, precision_use,]\n",
        "jaccard_scores = [jaccard_score_w2v, jaccard_score_use]\n",
        "result = pd.DataFrame(\n",
        "    {\n",
        "        'average_precision_scores' : avg_precision_scores,\n",
        "        'jaccard_scores' : jaccard_scores\n",
        "    }, index = ['Word2Vec','USE']\n",
        ")\n",
        "result.to_csv(f\"{output_dir}result_supervised2.csv\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 92,
      "metadata": {},
      "outputs": [
        {
          "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>average_precision_scores</th>\n",
              "      <th>jaccard_scores</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Word2Vec</th>\n",
              "      <td>0.085256</td>\n",
              "      <td>0.112318</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>USE</th>\n",
              "      <td>0.463946</td>\n",
              "      <td>0.500000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          average_precision_scores  jaccard_scores\n",
              "Word2Vec                  0.085256        0.112318\n",
              "USE                       0.463946        0.500000"
            ]
          },
          "execution_count": 92,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "result"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "authorship_tag": "ABX9TyMRwjeuBeVraeS78MTKQby9",
      "include_colab_link": true,
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.8.10"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}