Hybrid-MetaFeatures.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "\n",
    "np.random.seed(1)\n",
    "tf.random.set_seed(2)\n",
    "\n",
    "import pandas as pd\n",
    "import keras\n",
    "# from tqdm import tqdm\n",
    "from tqdm import tqdm_notebook as tqdm\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from sklearn.utils import class_weight\n",
    "from sklearn.metrics import f1_score, classification_report, log_loss\n",
    "\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "from keras.preprocessing.sequence import pad_sequences\n",
    "from keras.utils import to_categorical\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Embedding, LSTM, Dense, SpatialDropout1D, Bidirectional, Flatten\n",
    "from keras.layers import Dropout, Conv1D, GlobalMaxPool1D, GRU, GlobalAvgPool1D\n",
    "from keras.optimizers import Adam\n",
    "from keras.callbacks import ModelCheckpoint, EarlyStopping\n",
    "\n",
    "# print(os.listdir('./ML/LSTM/Binary'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['A1', 'A2', 'A3', 'hour', 'weekday', 'truth', 'month', 'age', 'sex',\n",
       "       'org', 'RMV', 'REM', 'RGLAD', 'RLFC', 'text'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('PA-B-LSTM.csv')\n",
    "data= data[data['hour']!='dummy']\n",
    "data['A1']=data['A1'].astype(\"category\")\n",
    "data['A2']=data['A2'].astype(\"category\")\n",
    "data['A3']=data['A3'].astype(\"category\")\n",
    "data['sex']=data['sex'].astype(\"category\")\n",
    "data['age']=data['age'].astype(\"category\")\n",
    "data['org']=data['org'].astype(\"category\")\n",
    "data['weekday']=data['weekday'].astype(\"category\")\n",
    "data['RMV']=data['RMV'].astype(\"category\")\n",
    "data['hour']=data['hour'].astype(\"int\")\n",
    "data['month']=data['month'].astype(\"int\")\n",
    "data['truth']=data['truth'].astype(\"category\")\n",
    "data.head()\n",
    "\n",
    "data.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>hour</th>\n",
       "      <th>truth</th>\n",
       "      <th>text</th>\n",
       "      <th>A1_0</th>\n",
       "      <th>A1_1</th>\n",
       "      <th>A2_0</th>\n",
       "      <th>A2_1</th>\n",
       "      <th>A3_0</th>\n",
       "      <th>A3_1</th>\n",
       "      <th>weekday_Fri</th>\n",
       "      <th>...</th>\n",
       "      <th>org_is-org</th>\n",
       "      <th>org_non-org</th>\n",
       "      <th>RMV_0</th>\n",
       "      <th>RMV_1</th>\n",
       "      <th>REM_0</th>\n",
       "      <th>REM_1</th>\n",
       "      <th>RGLAD_0</th>\n",
       "      <th>RGLAD_1</th>\n",
       "      <th>RLFC_0</th>\n",
       "      <th>RLFC_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>19</td>\n",
       "      <td>1</td>\n",
       "      <td>A lovely hike up Mt. McIntyre today...December...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>17</td>\n",
       "      <td>1</td>\n",
       "      <td>A lovely hike at the Research Forest Near Whit...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>19</td>\n",
       "      <td>1</td>\n",
       "      <td>A Lovely day for a hike in the Yukon! The leav...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "      <td>After yesterday’s massive snowstorm and Bully ...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>22</td>\n",
       "      <td>0</td>\n",
       "      <td>Me when my lil cousins try to wrestle me on Tu...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 44 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   hour truth                                               text  A1_0  A1_1  \\\n",
       "0    19     1  A lovely hike up Mt. McIntyre today...December...     0     1   \n",
       "1    17     1  A lovely hike at the Research Forest Near Whit...     0     1   \n",
       "2    19     1  A Lovely day for a hike in the Yukon! The leav...     0     1   \n",
       "3    21     0  After yesterday’s massive snowstorm and Bully ...     1     0   \n",
       "4    22     0  Me when my lil cousins try to wrestle me on Tu...     1     0   \n",
       "\n",
       "   A2_0  A2_1  A3_0  A3_1  weekday_Fri  ...  org_is-org  org_non-org  RMV_0  \\\n",
       "0     1     0     0     1            0  ...           0            1      0   \n",
       "1     1     0     1     0            0  ...           0            1      1   \n",
       "2     1     0     0     1            0  ...           0            1      0   \n",
       "3     1     0     1     0            1  ...           1            0      1   \n",
       "4     1     0     0     1            0  ...           0            1      1   \n",
       "\n",
       "   RMV_1  REM_0  REM_1  RGLAD_0  RGLAD_1  RLFC_0  RLFC_1  \n",
       "0      1      0      1        0        1       0       1  \n",
       "1      0      1      0        1        0       1       0  \n",
       "2      1      0      1        0        1       0       1  \n",
       "3      0      1      0        1        0       1       0  \n",
       "4      0      1      0        1        0       1       0  \n",
       "\n",
       "[5 rows x 44 columns]"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "data = pd.get_dummies(data, prefix=['A1', 'A2', 'A3', 'weekday', 'month', 'age', 'sex',\n",
    "       'org', 'RMV', 'REM', 'RGLAD', 'RLFC'], columns=['A1', 'A2', 'A3', 'weekday', 'month', 'age', 'sex',\n",
    "       'org', 'RMV', 'REM', 'RGLAD', 'RLFC'], drop_first=False)\n",
    "data.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = data.drop('truth', axis=1)\n",
    "\n",
    "y = data['truth']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "from keras.utils import np_utils\n",
    "from keras.utils import to_categorical\n",
    "\n",
    "# label_encoder object knows how to understand word labels.\n",
    "label_encoder = preprocessing.LabelEncoder()\n",
    "\n",
    "# Encode labels in column 'species'.\n",
    "y = label_encoder.fit_transform(y)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=42)\n",
    "\n",
    "from keras.utils import to_categorical\n",
    "y_train = to_categorical(y_train)\n",
    "y_test = to_categorical(y_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "dic = {\"`\":\"'\", \"’\":\"'\", \"‘\":\"'\", \"´\":\"'\",\"ain't\": \"is not\", \"aren't\": \"are not\",\"can't\": \"cannot\", \"'cause\": \"because\", \"could've\": \"could have\",\n",
    "                       \"couldn't\": \"could not\", \"didn't\": \"did not\",  \"doesn't\": \"does not\", \"don't\": \"do not\", \"hadn't\": \"had not\",\n",
    "                       \"hasn't\": \"has not\", \"haven't\": \"have not\", \"he'd\": \"he would\",\"he'll\": \"he will\", \"he's\": \"he is\", \"how'd\": \"how did\",\n",
    "                       \"how'd'y\": \"how do you\", \"how'll\": \"how will\", \"how's\": \"how is\",  \"I'd\": \"I would\", \"I'd've\": \"I would have\",\n",
    "                       \"I'll\": \"I will\", \"I'll've\": \"I will have\",\"I'm\": \"I am\", \"I've\": \"I have\", \"i'd\": \"i would\", \"i'd've\": \"i would have\",\n",
    "                       \"i'll\": \"i will\",  \"i'll've\": \"i will have\",\"i'm\": \"i am\", \"i've\": \"i have\", \"isn't\": \"is not\", \"it'd\": \"it would\",\n",
    "                       \"it'd've\": \"it would have\", \"it'll\": \"it will\", \"it'll've\": \"it will have\",\"it's\": \"it is\", \"let's\": \"let us\",\n",
    "                       \"ma'am\": \"madam\", \"mayn't\": \"may not\", \"might've\": \"might have\",\"mightn't\": \"might not\",\"mightn't've\": \"might not have\",\n",
    "                       \"must've\": \"must have\", \"mustn't\": \"must not\", \"mustn't've\": \"must not have\", \"needn't\": \"need not\",\n",
    "                       \"needn't've\": \"need not have\",\"o'clock\": \"of the clock\", \"oughtn't\": \"ought not\", \"oughtn't've\": \"ought not have\",\n",
    "                       \"shan't\": \"shall not\", \"sha'n't\": \"shall not\", \"shan't've\": \"shall not have\", \"she'd\": \"she would\",\n",
    "                       \"she'd've\": \"she would have\", \"she'll\": \"she will\", \"she'll've\": \"she will have\", \"she's\": \"she is\",\n",
    "                       \"should've\": \"should have\", \"shouldn't\": \"should not\", \"shouldn't've\": \"should not have\", \"so've\": \"so have\",\n",
    "                       \"so's\": \"so as\", \"this's\": \"this is\",\"that'd\": \"that would\", \"that'd've\": \"that would have\", \"that's\": \"that is\",\n",
    "                       \"there'd\": \"there would\", \"there'd've\": \"there would have\", \"there's\": \"there is\", \"here's\": \"here is\",\"they'd\": \"they would\",\n",
    "                       \"they'd've\": \"they would have\", \"they'll\": \"they will\", \"they'll've\": \"they will have\", \"they're\": \"they are\",\n",
    "                       \"they've\": \"they have\", \"to've\": \"to have\", \"wasn't\": \"was not\", \"we'd\": \"we would\", \"we'd've\": \"we would have\",\n",
    "                       \"we'll\": \"we will\", \"we'll've\": \"we will have\", \"we're\": \"we are\", \"we've\": \"we have\", \"weren't\": \"were not\",\n",
    "                       \"what'll\": \"what will\", \"what'll've\": \"what will have\", \"what're\": \"what are\",  \"what's\": \"what is\", \"what've\": \"what have\",\n",
    "                       \"when's\": \"when is\", \"when've\": \"when have\", \"where'd\": \"where did\", \"where's\": \"where is\", \"where've\": \"where have\",\n",
    "                       \"who'll\": \"who will\", \"who'll've\": \"who will have\", \"who's\": \"who is\", \"who've\": \"who have\", \"why's\": \"why is\",\n",
    "                       \"why've\": \"why have\", \"will've\": \"will have\", \"won't\": \"will not\", \"won't've\": \"will not have\", \"would've\": \"would have\",\n",
    "                       \"wouldn't\": \"would not\", \"wouldn't've\": \"would not have\", \"y'all\": \"you all\", \"y'all'd\": \"you all would\",\n",
    "                       \"y'all'd've\": \"you all would have\",\"y'all're\": \"you all are\",\"y'all've\": \"you all have\",\"you'd\": \"you would\",\n",
    "                       \"you'd've\": \"you would have\", \"you'll\": \"you will\", \"you'll've\": \"you will have\", \"you're\": \"you are\", \"you've\": \"you have\",\n",
    "                       \"It's\": 'It is', \"Can't\": 'Can not', \"I've\": 'I have', \"\\n\":\" \", \"—\":\"\", \".\":\"\", \"…\":\"\", \"!\":\"\", \":\":\" \",\n",
    "                       \"-\":\" \",\"•\":\"\"}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocess_text(sen):\n",
    "\n",
    "    # Remove punctuations and numbers\n",
    "    sentence = re.sub('[^a-zA-Z]', ' ', sen)\n",
    "\n",
    "    # Single character removal\n",
    "    sentence = re.sub(r\"\\s+[a-zA-Z]\\s+\", ' ', sentence)\n",
    "\n",
    "    # Removing multiple spaces\n",
    "    sentence = re.sub(r'\\s+', ' ', sentence)\n",
    "    for i, j in dic.items():\n",
    "        sentence = sentence.lower().replace(i, j)\n",
    "    sentence = re.sub(r\"http[s]?://t.co/[A-Za-z0-9]*\",\" \",sentence) #URLs\n",
    "\n",
    "    return sentence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "X1_train = []\n",
    "sentences = list(X_train[\"text\"])\n",
    "for sen in sentences:\n",
    "    X1_train.append(preprocess_text(sen))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "X1_test = []\n",
    "sentences = list(X_test[\"text\"])\n",
    "for sen in sentences:\n",
    "    X1_test.append(preprocess_text(sen))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokenizer = Tokenizer()\n",
    "tokenizer.fit_on_texts(X1_train)\n",
    "\n",
    "X1_train = tokenizer.texts_to_sequences(X1_train)\n",
    "X1_test = tokenizer.texts_to_sequences(X1_test)\n",
    "\n",
    "vocab_size = len(tokenizer.word_index) + 1\n",
    "\n",
    "maxlen = 27\n",
    "\n",
    "X1_train = pad_sequences(X1_train, padding='post', maxlen=maxlen)\n",
    "X1_test = pad_sequences(X1_test, padding='post', maxlen=maxlen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from numpy import array\n",
    "from numpy import asarray\n",
    "from numpy import zeros\n",
    "\n",
    "embeddings_dictionary = dict()\n",
    "\n",
    "# glove_file = open('glove.twitter.27B.200d.txt', encoding=\"utf8\")\n",
    "\n",
    "# for line in glove_file:\n",
    "#     records = line.split()\n",
    "#     word = records[0]\n",
    "#     vector_dimensions = asarray(records[1:], dtype='float32')\n",
    "#     embeddings_dictionary[word] = vector_dimensions\n",
    "\n",
    "# glove_file.close()\n",
    "\n",
    "embedding_matrix = zeros((vocab_size, 200))\n",
    "for word, index in tokenizer.word_index.items():\n",
    "    embedding_vector = embeddings_dictionary.get(word)\n",
    "    if embedding_vector is not None:\n",
    "        embedding_matrix[index] = embedding_vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['hour', 'text', 'A1_0', 'A1_1', 'A2_0', 'A2_1', 'A3_0', 'A3_1',\n",
       "       'weekday_Fri', 'weekday_Mon', 'weekday_Sat', 'weekday_Sun',\n",
       "       'weekday_Thu', 'weekday_Tue', 'weekday_Wed', 'month_1', 'month_2',\n",
       "       'month_3', 'month_4', 'month_5', 'month_6', 'month_7', 'month_8',\n",
       "       'month_9', 'month_10', 'month_11', 'month_12', 'age_19-29', 'age_30-39',\n",
       "       'age_<=18', 'age_>=40', 'sex_female', 'sex_male', 'org_is-org',\n",
       "       'org_non-org', 'RMV_0', 'RMV_1', 'REM_0', 'REM_1', 'RGLAD_0', 'RGLAD_1',\n",
       "       'RLFC_0', 'RLFC_1'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "X2_train = X_train[[ 'A1_0', 'A1_1', 'A2_0', 'A2_1', 'A3_0']].values\n",
    "X2_test = X_test[['A1_0', 'A1_1', 'A2_0', 'A2_1', 'A3_0']]\n",
    "ncols=X2_train.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.layers import Input\n",
    "from keras.layers import Dropout, Conv1D, GlobalMaxPool1D, GRU, GlobalAvgPool1D\n",
    "input_1 = Input(shape=(maxlen,))\n",
    "\n",
    "input_2 = Input(shape=(ncols,))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "embedding_layer = Embedding(vocab_size, 200, weights=[embedding_matrix], trainable=False)(input_1)\n",
    "#LSTM_Layer_1 = Bidirectional(LSTM(128, dropout=0.25, recurrent_dropout=0.25))(embedding_layer)\n",
    "LSTM_Layer_1 = LSTM(128, dropout=0.25, recurrent_dropout=0.25)(embedding_layer)\n",
    "LSTM_Layer_2 = Dense(64, activation='relu')(LSTM_Layer_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [],
   "source": [
    "dense_layer_1 = Dense(10, activation='relu')(input_2)\n",
    "dense_layer_2 = Dense(64, activation='relu')(dense_layer_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.layers import concatenate\n",
    "from keras.models import Model\n",
    "from keras.layers import Dropout, Conv1D, GlobalMaxPool1D, GRU, GlobalAvgPool1D\n",
    "\n",
    "concat_layer = concatenate([LSTM_Layer_2, dense_layer_2])\n",
    "dense_layer_3 = Dense(10, activation='relu')(concat_layer)\n",
    "output = Dense(1, activation='softmax')(dense_layer_3)\n",
    "model = Model(inputs=[input_1, input_2], outputs=output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"functional_7\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_7 (InputLayer)            [(None, 27)]         0                                            \n",
      "__________________________________________________________________________________________________\n",
      "embedding_3 (Embedding)         (None, 27, 200)      1928000     input_7[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "input_8 (InputLayer)            [(None, 5)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "lstm_3 (LSTM)                   (None, 128)          168448      embedding_3[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "dense_16 (Dense)                (None, 10)           60          input_8[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_15 (Dense)                (None, 64)           8256        lstm_3[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "dense_17 (Dense)                (None, 64)           704         dense_16[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 128)          0           dense_15[0][0]                   \n",
      "                                                                 dense_17[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "dense_18 (Dense)                (None, 10)           1290        concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dense_19 (Dense)                (None, 2)            22          dense_18[0][0]                   \n",
      "==================================================================================================\n",
      "Total params: 2,106,780\n",
      "Trainable params: 178,780\n",
      "Non-trainable params: 1,928,000\n",
      "__________________________________________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "from keras import backend as K\n",
    "import tensorflow as tf\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "\n",
    "def recall_m(y_true, y_pred):\n",
    "    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
    "    possible_positives = K.sum(K.round(K.clip(y_true, 0, 1)))\n",
    "    recall = true_positives / (possible_positives + K.epsilon())\n",
    "    return recall\n",
    "\n",
    "def precision_m(y_true, y_pred):\n",
    "    true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
    "    predicted_positives = K.sum(K.round(K.clip(y_pred, 0, 1)))\n",
    "    precision = true_positives / (predicted_positives + K.epsilon())\n",
    "    return precision\n",
    "\n",
    "def f1_m(y_true, y_pred):\n",
    "    precision = precision_m(y_true, y_pred)\n",
    "    recall = recall_m(y_true, y_pred)\n",
    "    return 2*((precision*recall)/(precision+recall+K.epsilon()))\n",
    "\n",
    "def auroc(y_true, y_pred):\n",
    "    return tf.py_function(roc_auc_score, (y_true, y_pred), tf.double)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['acc', f1_m,precision_m, recall_m])\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from keras.utils import plot_model\n",
    "plot_model(model, to_file='model_plot3.png', show_shapes=True, show_layer_names=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "14/14 [==============================] - 1s 105ms/step - loss: 0.7086 - acc: 0.4110 - f1_m: 0.4104 - precision_m: 0.4104 - recall_m: 0.4104 - val_loss: 0.6870 - val_acc: 0.7512 - val_f1_m: 0.7586 - val_precision_m: 0.7586 - val_recall_m: 0.7586\n",
      "Epoch 2/50\n",
      "14/14 [==============================] - 1s 80ms/step - loss: 0.6770 - acc: 0.7908 - f1_m: 0.7921 - precision_m: 0.7921 - recall_m: 0.7921 - val_loss: 0.6709 - val_acc: 0.7719 - val_f1_m: 0.7792 - val_precision_m: 0.7792 - val_recall_m: 0.7792\n",
      "Epoch 3/50\n",
      "14/14 [==============================] - 1s 85ms/step - loss: 0.6640 - acc: 0.7844 - f1_m: 0.7855 - precision_m: 0.7855 - recall_m: 0.7855 - val_loss: 0.6586 - val_acc: 0.7604 - val_f1_m: 0.7725 - val_precision_m: 0.7725 - val_recall_m: 0.7725\n",
      "Epoch 4/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.6488 - acc: 0.7781 - f1_m: 0.7771 - precision_m: 0.7771 - recall_m: 0.7771 - val_loss: 0.6414 - val_acc: 0.7880 - val_f1_m: 0.7959 - val_precision_m: 0.7959 - val_recall_m: 0.7959\n",
      "Epoch 5/50\n",
      "14/14 [==============================] - 1s 81ms/step - loss: 0.6267 - acc: 0.8017 - f1_m: 0.8009 - precision_m: 0.8009 - recall_m: 0.8009 - val_loss: 0.6197 - val_acc: 0.7880 - val_f1_m: 0.7959 - val_precision_m: 0.7959 - val_recall_m: 0.7959\n",
      "Epoch 6/50\n",
      "14/14 [==============================] - 1s 80ms/step - loss: 0.6011 - acc: 0.8017 - f1_m: 0.8009 - precision_m: 0.8009 - recall_m: 0.8009 - val_loss: 0.5958 - val_acc: 0.7880 - val_f1_m: 0.7959 - val_precision_m: 0.7959 - val_recall_m: 0.7959\n",
      "Epoch 7/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.5745 - acc: 0.8052 - f1_m: 0.8069 - precision_m: 0.8069 - recall_m: 0.8069 - val_loss: 0.5687 - val_acc: 0.8502 - val_f1_m: 0.8578 - val_precision_m: 0.8578 - val_recall_m: 0.8578\n",
      "Epoch 8/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.5491 - acc: 0.8530 - f1_m: 0.8528 - precision_m: 0.8528 - recall_m: 0.8528 - val_loss: 0.5470 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 9/50\n",
      "14/14 [==============================] - 1s 86ms/step - loss: 0.5254 - acc: 0.8588 - f1_m: 0.8592 - precision_m: 0.8592 - recall_m: 0.8592 - val_loss: 0.5266 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 10/50\n",
      "14/14 [==============================] - 1s 81ms/step - loss: 0.5054 - acc: 0.8588 - f1_m: 0.8610 - precision_m: 0.8610 - recall_m: 0.8610 - val_loss: 0.5144 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 11/50\n",
      "14/14 [==============================] - 1s 81ms/step - loss: 0.4931 - acc: 0.8588 - f1_m: 0.8597 - precision_m: 0.8597 - recall_m: 0.8597 - val_loss: 0.5060 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 12/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.4858 - acc: 0.8588 - f1_m: 0.8601 - precision_m: 0.8601 - recall_m: 0.8601 - val_loss: 0.4999 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 13/50\n",
      "14/14 [==============================] - 1s 81ms/step - loss: 0.4794 - acc: 0.8588 - f1_m: 0.8588 - precision_m: 0.8588 - recall_m: 0.8588 - val_loss: 0.4955 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 14/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.4735 - acc: 0.8588 - f1_m: 0.8588 - precision_m: 0.8588 - recall_m: 0.8588 - val_loss: 0.4898 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 15/50\n",
      "14/14 [==============================] - 1s 80ms/step - loss: 0.4688 - acc: 0.8588 - f1_m: 0.8606 - precision_m: 0.8606 - recall_m: 0.8606 - val_loss: 0.4852 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 16/50\n",
      "14/14 [==============================] - 1s 77ms/step - loss: 0.4642 - acc: 0.8588 - f1_m: 0.8588 - precision_m: 0.8588 - recall_m: 0.8588 - val_loss: 0.4808 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 17/50\n",
      "14/14 [==============================] - 1s 83ms/step - loss: 0.4602 - acc: 0.8588 - f1_m: 0.8575 - precision_m: 0.8575 - recall_m: 0.8575 - val_loss: 0.4772 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 18/50\n",
      "14/14 [==============================] - 1s 86ms/step - loss: 0.4566 - acc: 0.8588 - f1_m: 0.8579 - precision_m: 0.8579 - recall_m: 0.8579 - val_loss: 0.4721 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 19/50\n",
      "14/14 [==============================] - 1s 88ms/step - loss: 0.4528 - acc: 0.8588 - f1_m: 0.8566 - precision_m: 0.8566 - recall_m: 0.8566 - val_loss: 0.4695 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 20/50\n",
      "14/14 [==============================] - 1s 76ms/step - loss: 0.4495 - acc: 0.8588 - f1_m: 0.8592 - precision_m: 0.8592 - recall_m: 0.8592 - val_loss: 0.4653 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 21/50\n",
      "14/14 [==============================] - 1s 78ms/step - loss: 0.4461 - acc: 0.8588 - f1_m: 0.8575 - precision_m: 0.8575 - recall_m: 0.8575 - val_loss: 0.4615 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 22/50\n",
      "14/14 [==============================] - 1s 73ms/step - loss: 0.4432 - acc: 0.8588 - f1_m: 0.8597 - precision_m: 0.8597 - recall_m: 0.8597 - val_loss: 0.4582 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 23/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.4403 - acc: 0.8588 - f1_m: 0.8588 - precision_m: 0.8588 - recall_m: 0.8588 - val_loss: 0.4550 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 24/50\n",
      "14/14 [==============================] - 1s 91ms/step - loss: 0.4376 - acc: 0.8588 - f1_m: 0.8575 - precision_m: 0.8575 - recall_m: 0.8575 - val_loss: 0.4529 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 25/50\n",
      "14/14 [==============================] - 1s 83ms/step - loss: 0.4352 - acc: 0.8588 - f1_m: 0.8579 - precision_m: 0.8579 - recall_m: 0.8579 - val_loss: 0.4506 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 26/50\n",
      "14/14 [==============================] - 1s 77ms/step - loss: 0.4328 - acc: 0.8588 - f1_m: 0.8606 - precision_m: 0.8606 - recall_m: 0.8606 - val_loss: 0.4467 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 27/50\n",
      "14/14 [==============================] - 1s 76ms/step - loss: 0.4304 - acc: 0.8588 - f1_m: 0.8579 - precision_m: 0.8579 - recall_m: 0.8579 - val_loss: 0.4467 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 28/50\n",
      "14/14 [==============================] - 1s 70ms/step - loss: 0.4278 - acc: 0.8588 - f1_m: 0.8592 - precision_m: 0.8592 - recall_m: 0.8592 - val_loss: 0.4423 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 29/50\n",
      "14/14 [==============================] - 1s 71ms/step - loss: 0.4259 - acc: 0.8588 - f1_m: 0.8579 - precision_m: 0.8579 - recall_m: 0.8579 - val_loss: 0.4438 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 30/50\n",
      "14/14 [==============================] - 1s 78ms/step - loss: 0.4236 - acc: 0.8588 - f1_m: 0.8592 - precision_m: 0.8592 - recall_m: 0.8592 - val_loss: 0.4400 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 31/50\n",
      "14/14 [==============================] - 1s 76ms/step - loss: 0.4217 - acc: 0.8588 - f1_m: 0.8579 - precision_m: 0.8579 - recall_m: 0.8579 - val_loss: 0.4361 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 32/50\n",
      "14/14 [==============================] - 1s 74ms/step - loss: 0.4199 - acc: 0.8588 - f1_m: 0.8597 - precision_m: 0.8597 - recall_m: 0.8597 - val_loss: 0.4367 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 33/50\n",
      "14/14 [==============================] - 1s 76ms/step - loss: 0.4175 - acc: 0.8588 - f1_m: 0.8606 - precision_m: 0.8606 - recall_m: 0.8606 - val_loss: 0.4341 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 34/50\n",
      "14/14 [==============================] - 1s 74ms/step - loss: 0.4150 - acc: 0.8588 - f1_m: 0.8610 - precision_m: 0.8610 - recall_m: 0.8610 - val_loss: 0.4316 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 35/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.4135 - acc: 0.8588 - f1_m: 0.8597 - precision_m: 0.8597 - recall_m: 0.8597 - val_loss: 0.4301 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 36/50\n",
      "14/14 [==============================] - 1s 72ms/step - loss: 0.4113 - acc: 0.8588 - f1_m: 0.8592 - precision_m: 0.8592 - recall_m: 0.8592 - val_loss: 0.4266 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 37/50\n",
      "14/14 [==============================] - 1s 70ms/step - loss: 0.4095 - acc: 0.8588 - f1_m: 0.8606 - precision_m: 0.8606 - recall_m: 0.8606 - val_loss: 0.4286 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 38/50\n",
      "14/14 [==============================] - 1s 72ms/step - loss: 0.4079 - acc: 0.8588 - f1_m: 0.8597 - precision_m: 0.8597 - recall_m: 0.8597 - val_loss: 0.4245 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 39/50\n",
      "14/14 [==============================] - 1s 73ms/step - loss: 0.4058 - acc: 0.8588 - f1_m: 0.8584 - precision_m: 0.8584 - recall_m: 0.8584 - val_loss: 0.4224 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 40/50\n",
      "14/14 [==============================] - 1s 74ms/step - loss: 0.4047 - acc: 0.8588 - f1_m: 0.8610 - precision_m: 0.8610 - recall_m: 0.8610 - val_loss: 0.4207 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 41/50\n",
      "14/14 [==============================] - 1s 76ms/step - loss: 0.4029 - acc: 0.8588 - f1_m: 0.8592 - precision_m: 0.8592 - recall_m: 0.8592 - val_loss: 0.4210 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 42/50\n",
      "14/14 [==============================] - 1s 71ms/step - loss: 0.4015 - acc: 0.8588 - f1_m: 0.8606 - precision_m: 0.8606 - recall_m: 0.8606 - val_loss: 0.4164 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 43/50\n",
      "14/14 [==============================] - 1s 72ms/step - loss: 0.4002 - acc: 0.8588 - f1_m: 0.8588 - precision_m: 0.8588 - recall_m: 0.8588 - val_loss: 0.4147 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 44/50\n",
      "14/14 [==============================] - 1s 74ms/step - loss: 0.3979 - acc: 0.8588 - f1_m: 0.8570 - precision_m: 0.8570 - recall_m: 0.8570 - val_loss: 0.4118 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 45/50\n",
      "14/14 [==============================] - 1s 82ms/step - loss: 0.3965 - acc: 0.8588 - f1_m: 0.8575 - precision_m: 0.8575 - recall_m: 0.8575 - val_loss: 0.4111 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 46/50\n",
      "14/14 [==============================] - 1s 70ms/step - loss: 0.3953 - acc: 0.8588 - f1_m: 0.8601 - precision_m: 0.8601 - recall_m: 0.8601 - val_loss: 0.4134 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 47/50\n",
      "14/14 [==============================] - 1s 73ms/step - loss: 0.3938 - acc: 0.8588 - f1_m: 0.8588 - precision_m: 0.8588 - recall_m: 0.8588 - val_loss: 0.4079 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 48/50\n",
      "14/14 [==============================] - 1s 72ms/step - loss: 0.3922 - acc: 0.8588 - f1_m: 0.8584 - precision_m: 0.8584 - recall_m: 0.8584 - val_loss: 0.4069 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 49/50\n",
      "14/14 [==============================] - 1s 74ms/step - loss: 0.3908 - acc: 0.8588 - f1_m: 0.8570 - precision_m: 0.8570 - recall_m: 0.8570 - val_loss: 0.4052 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n",
      "Epoch 50/50\n",
      "14/14 [==============================] - 1s 74ms/step - loss: 0.3898 - acc: 0.8588 - f1_m: 0.8610 - precision_m: 0.8610 - recall_m: 0.8610 - val_loss: 0.4045 - val_acc: 0.8618 - val_f1_m: 0.8645 - val_precision_m: 0.8645 - val_recall_m: 0.8645\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(x=[X1_train, X2_train], y=y_train, batch_size=128, epochs=50, verbose=1, validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30/30 [==============================] - 0s 6ms/step - loss: 0.4327 - acc: 0.8344 - f1_m: 0.8396 - precision_m: 0.8396 - recall_m: 0.8396\n"
     ]
    }
   ],
   "source": [
    "loss, accuracy, f1_score, precision, recall  = model.evaluate(x=[X1_test, X2_test], y=y_test, verbose=1)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(history.history['acc'])\n",
    "plt.plot(history.history['val_acc'])\n",
    "\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train','test'], loc='upper left')\n",
    "plt.show()\n",
    "\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train','test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Automatically created module for IPython interactive environment\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(__doc__)\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn import svm, datasets\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "from sklearn.metrics import average_precision_score\n",
    "from sklearn.preprocessing import label_binarize\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "predictions = model.predict([X1_test, X2_test])\n",
    "y_pred = (predictions > 0.7)\n",
    "y_test = label_binarize(y_test, classes=['YY','YN'])\n",
    "y_pred = label_binarize(y_pred, classes=['YY','YN'])\n",
    "n_classes = 2\n",
    "\n",
    "\n",
    "\n",
    "# Compute Precision-Recall and plot curve\n",
    "precision = dict()\n",
    "recall = dict()\n",
    "average_precision = dict()\n",
    "for i in range(n_classes):\n",
    "    precision[i], recall[i], _ = precision_recall_curve(y_test[:, i],\n",
    "                                                        y_pred[:, i])\n",
    "    average_precision[i] = average_precision_score(y_test[:, i], y_pred[:, i])\n",
    "    \n",
    "\n",
    "# Compute micro-average ROC curve and ROC area\n",
    "precision[\"micro\"], recall[\"micro\"], _ = precision_recall_curve(y_test.ravel(),\n",
    "    y_pred.ravel())\n",
    "average_precision[\"micro\"] = average_precision_score(y_test, y_pred,\n",
    "                                                     average=\"micro\")\n",
    "\n",
    "  \n",
    "# Plot Precision-Recall curve\n",
    "plt.clf()\n",
    "plt.plot(recall[0], precision[0], label='Precision-Recall curve')\n",
    "plt.xlabel('Recall')\n",
    "plt.ylabel('Precision')\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.title('Precision-Recall example: AUC={0:0.2f}'.format(average_precision[0]))\n",
    "plt.legend(loc=\"lower left\")\n",
    "plt.show()\n",
    "\n",
    "# Plot Precision-Recall curve for each class\n",
    "plt.clf()\n",
    "plt.plot(recall[\"micro\"], precision[\"micro\"],\n",
    "         label='micro-average Precision-recall curve (area = {0:0.2f})'\n",
    "               ''.format(average_precision[\"micro\"]))\n",
    "for i in range(n_classes):\n",
    "    plt.plot(recall[i], precision[i],\n",
    "             label='Precision-recall curve of class {0} (area = {1:0.2f})'\n",
    "                   ''.format(i, average_precision[i]))\n",
    "\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('Recall')\n",
    "plt.ylabel('Precision')\n",
    "plt.title('Extension of Precision-Recall curve to multi-class')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}