analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !pip install -q -U tensorflow_hub"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import tensorflow_hub as hub\n",
    "from keras.callbacks import EarlyStopping\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "%config InlineBackend.figure_format='retina'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### select selected columns from the pandas dataframe\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('wine-reviews.csv', usecols=['country', 'description',  'points', 'price', 'variety', 'winery' ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><div id=1126cf22-1c0d-4578-8a41-20ea77d6a14a style=\"display:none; background-color:#9D6CFF; color:white; width:200px; height:30px; padding-left:5px; border-radius:4px; flex-direction:row; justify-content:space-around; align-items:center;\" onmouseover=\"this.style.backgroundColor='#BA9BF8'\" onmouseout=\"this.style.backgroundColor='#9D6CFF'\" onclick=\"window.commands?.execute('create-mitosheet-from-dataframe-output');\">See Full Dataframe in Mito</div> <script> if (window.commands.hasCommand('create-mitosheet-from-dataframe-output')) document.getElementById('1126cf22-1c0d-4578-8a41-20ea77d6a14a').style.display = 'flex' </script> <table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>country</th>\n",
       "      <th>description</th>\n",
       "      <th>points</th>\n",
       "      <th>price</th>\n",
       "      <th>variety</th>\n",
       "      <th>winery</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Italy</td>\n",
       "      <td>Aromas include tropical fruit, broom, brimstone and dried herb. The palate isn't overly expressive, offering unripened apple, citrus and dried sage alongside brisk acidity.</td>\n",
       "      <td>87</td>\n",
       "      <td>NaN</td>\n",
       "      <td>White Blend</td>\n",
       "      <td>Nicosia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Portugal</td>\n",
       "      <td>This is ripe and fruity, a wine that is smooth while still structured. Firm tannins are filled out with juicy red berry fruits and freshened with acidity. It's  already drinkable, although it will certainly be better from 2016.</td>\n",
       "      <td>87</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Portuguese Red</td>\n",
       "      <td>Quinta dos Avidagos</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>US</td>\n",
       "      <td>Tart and snappy, the flavors of lime flesh and rind dominate. Some green pineapple pokes through, with crisp acidity underscoring the flavors. The wine was all stainless-steel fermented.</td>\n",
       "      <td>87</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Pinot Gris</td>\n",
       "      <td>Rainstorm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>US</td>\n",
       "      <td>Pineapple rind, lemon pith and orange blossom start off the aromas. The palate is a bit more opulent, with notes of honey-drizzled guava and mango giving way to a slightly astringent, semidry finish.</td>\n",
       "      <td>87</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Riesling</td>\n",
       "      <td>St. Julian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>US</td>\n",
       "      <td>Much like the regular bottling from 2012, this comes across as rather rough and tannic, with rustic, earthy, herbal characteristics. Nonetheless, if you think of it as a pleasantly unfussy country wine, it's a good companion to a hearty winter stew.</td>\n",
       "      <td>87</td>\n",
       "      <td>65.0</td>\n",
       "      <td>Pinot Noir</td>\n",
       "      <td>Sweet Cheeks</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table></div>"
      ],
      "text/plain": [
       "    country                                        description  points  price  \\\n",
       "0     Italy  Aromas include tropical fruit, broom, brimston...      87    NaN   \n",
       "1  Portugal  This is ripe and fruity, a wine that is smooth...      87   15.0   \n",
       "2        US  Tart and snappy, the flavors of lime flesh and...      87   14.0   \n",
       "3        US  Pineapple rind, lemon pith and orange blossom ...      87   13.0   \n",
       "4        US  Much like the regular bottling from 2012, this...      87   65.0   \n",
       "\n",
       "          variety               winery  \n",
       "0     White Blend              Nicosia  \n",
       "1  Portuguese Red  Quinta dos Avidagos  \n",
       "2      Pinot Gris            Rainstorm  \n",
       "3        Riesling           St. Julian  \n",
       "4      Pinot Noir         Sweet Cheeks  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### dropna values from the description and the points columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.dropna(subset=['description', 'points'], inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> This is the summary for the final df </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "country          63\n",
       "description       0\n",
       "points            0\n",
       "price          8996\n",
       "variety           1\n",
       "winery            0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 129971 entries, 0 to 129970\n",
      "Data columns (total 6 columns):\n",
      " #   Column       Non-Null Count   Dtype  \n",
      "---  ------       --------------   -----  \n",
      " 0   country      129908 non-null  object \n",
      " 1   description  129971 non-null  object \n",
      " 2   points       129971 non-null  int64  \n",
      " 3   price        120975 non-null  float64\n",
      " 4   variety      129970 non-null  object \n",
      " 5   winery       129971 non-null  object \n",
      "dtypes: float64(1), int64(1), object(4)\n",
      "memory usage: 6.9+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 261,
       "width": 401
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(df.points, bins=15)\n",
    "plt.xlabel('Points')\n",
    "plt.ylabel('Distribution')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Apply lambda function tp perticular column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['label'] = df.points.apply(lambda x: 1 if x >= 90 else 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    80926\n",
       "1    49045\n",
       "Name: label, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.label.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# we can also use the following code to create a new column called 'label' instead of using the apply function\n",
    "# df['label'] = (df['points'] >= 90).astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[['description', 'label']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "train, val, test = np.split(df.sample(frac=1), [int(.8*len(df)), int(.9*len(df))])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(103976, 2) (12997, 2) (12998, 2)\n"
     ]
    }
   ],
   "source": [
    "print(train.shape, val.shape, test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### What is the puropose of the following code?\n",
    "### look it into details later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "def df_to_dataset(dataframe, shuffle=True, batch_size=1024):\n",
    "  df = dataframe.copy()\n",
    "  labels = df.pop('label')\n",
    "  df = df[\"description\"]\n",
    "  ds = tf.data.Dataset.from_tensor_slices((df, labels))\n",
    "  if shuffle:\n",
    "    ds = ds.shuffle(buffer_size=len(dataframe))\n",
    "  ds = ds.batch(batch_size)\n",
    "  ds = ds.prefetch(tf.data.AUTOTUNE)\n",
    "  return ds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = df_to_dataset(train)\n",
    "valid_data = df_to_dataset(val)\n",
    "test_data = df_to_dataset(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(<tf.Tensor: shape=(1024,), dtype=string, numpy=\n",
      "array([b'A very rich, dramatic, indisputably Napa Cabernet. The fruit is ripe in blackberries and cassis, the tannin structure is vibrant and living, and the oak application is superb. A fashionable Cab that should develop over the next eight years.',\n",
      "       b'This is a pleasant red wine from Piedmont in northern Italy made with Nebbiolo and Vespolina grapes for a clean delivery of berry, spice and cola aromas. The wine is tonic and fresh in the mouth with clean acidity and playful bite of exotic spice on the close.',\n",
      "       b'This is a broad, sweet, full-bodied dessert wine that lacks the crisp acids necessary for a higher degree of balance. Instead, it delivers intense honey and pineapple flavors that stick stubbornly to the palate. Probably best with sweet desserts.',\n",
      "       ...,\n",
      "       b'Touched by spice from wood aging, this is a balanced wine. Its toastiness supports the ripe peach and apricot fruits, lifted by acidity. It will be a good wine; drink from 2018.',\n",
      "       b'Fresh and fragrant, this is a lively, bright and tangy wine. With crisp acidity and raspberry fruit that is twisted with orange peel it is ready to drink.',\n",
      "       b\"The Matariki style of late harvest wine is to avoid overly sugary renditions in favor of balance. In this, the 2000 succeeds admirably, combining bergamot, pineapple and apricot flavors in a lightly sweet yet tart wine. Best with fresh fruit and cheeses, as it's not rich enough to match with anything sweeter.\"],\n",
      "      dtype=object)>, <tf.Tensor: shape=(1024,), dtype=int64, numpy=array([1, 0, 0, ..., 0, 0, 0])>)\n"
     ]
    }
   ],
   "source": [
    "for i in train_data.take(1):\n",
    "  print(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-06-21 00:10:43.039473: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n"
     ]
    }
   ],
   "source": [
    "embedding = \"https://tfhub.dev/google/nnlm-en-dim50/2\"\n",
    "hub_layer = hub.KerasLayer(embedding, dtype=tf.string, trainable=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: shape=(1024,), dtype=string, numpy=\n",
       "array([b'Completely stainless-fermented and aged, this is a blockbuster of a white wine, refreshing and exotic in floral intoxication. Dry and light-bodied, it offers an edginess of slate-like minerality along with aromas of apricot and peach, finishing in a dusting of white pepper.',\n",
       "       b'The label says Napa Valley but the grapes come from Carneros, and you can taste the cool fog in the crisp acidity that frames long hangtime flavors of ripe pineapple, guava, nectarine, roasted peach and lime custard. This is a pretty Chard, balanced and dry, and it impresses for its volume and length.',\n",
       "       b'Ripe white grapefruit and lychee notes perfume this refreshing, easy-drinking Gr\\xc3\\xbcner Veltliner. On the palate, gooseberry and lime flavors are delicate in concentration but it finishes with an invigorating brace of crushed minerals.',\n",
       "       ...,\n",
       "       b'This blend of Sangiovese and Merlot starts with aromas of espresso, toasted oak and plum. The palate offers one-dimensional plum and cherry with oak notes alongside firm tannins and uplifting acidity.',\n",
       "       b'A direct reflection of the 2012 vintage, with notes of wild strawberry, sarsaparilla, clove and cinnamon. The flavors are medium bodied in style, juicy and appealing.',\n",
       "       b'A generous wine, this is packed with red fruits as well as firmly structured. It conveys the fresh acidity of the year, with fine smoky tannins. It will be an impressive wine. Drink from 2020.'],\n",
       "      dtype=object)>"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(train_data)[0][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tf.Tensor: shape=(1024, 50), dtype=float32, numpy=\n",
       "array([[-0.0012439 , -0.09365706, -0.12321281, ..., -0.3201542 ,\n",
       "        -0.1284347 ,  0.05955575],\n",
       "       [ 0.33945358, -0.23618123, -0.2305099 , ..., -0.08266268,\n",
       "         0.27211103,  0.15572451],\n",
       "       [ 0.28139442, -0.12988901,  0.09559803, ...,  0.18410209,\n",
       "        -0.05989112, -0.11929152],\n",
       "       ...,\n",
       "       [ 0.45168313, -0.13605252, -0.31841782, ..., -0.10140887,\n",
       "        -0.1788628 ,  0.2342917 ],\n",
       "       [ 0.45715016, -0.16173916, -0.12438109, ...,  0.10529581,\n",
       "        -0.17982656,  0.04192869],\n",
       "       [ 0.08921701, -0.22221851, -0.2257588 , ...,  0.0052031 ,\n",
       "         0.11706206, -0.09028736]], dtype=float32)>"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hub_layer(list(train_data)[0][0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define a deep neural net model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_model():\n",
    "    model = tf.keras.Sequential()\n",
    "    model.add(hub_layer)\n",
    "    model.add(tf.keras.layers.Dense(16, activation='relu'))\n",
    "    model.add(tf.keras.layers.Dense(16, activation='relu'))\n",
    "    model.add(tf.keras.layers.Dense(1, activation='sigmoid'))\n",
    "    model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),\n",
    "              loss=tf.keras.losses.BinaryCrossentropy(),\n",
    "              metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = get_model()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### This is model evaluation on test data and validation data without training the network and accuracy is pretty bad 38.22%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "102/102 [==============================] - 2s 21ms/step - loss: 0.7354 - accuracy: 0.3822\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.7354397177696228, 0.3821747303009033]"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(train_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13/13 [==============================] - 0s 5ms/step - loss: 0.7118 - accuracy: 0.3856\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.7117805480957031, 0.3856274485588074]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(valid_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Training the model using 10 epochs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "102/102 [==============================] - 15s 150ms/step - loss: 0.2529 - accuracy: 0.8935 - val_loss: 0.4000 - val_accuracy: 0.8287\n",
      "Epoch 2/10\n",
      "102/102 [==============================] - 16s 152ms/step - loss: 0.2458 - accuracy: 0.8962 - val_loss: 0.4008 - val_accuracy: 0.8291\n",
      "Epoch 3/10\n",
      "102/102 [==============================] - 16s 154ms/step - loss: 0.2365 - accuracy: 0.9007 - val_loss: 0.4053 - val_accuracy: 0.8283\n",
      "Epoch 4/10\n",
      "102/102 [==============================] - 15s 150ms/step - loss: 0.2276 - accuracy: 0.9049 - val_loss: 0.4120 - val_accuracy: 0.8300\n",
      "Epoch 5/10\n",
      "102/102 [==============================] - 16s 152ms/step - loss: 0.2183 - accuracy: 0.9090 - val_loss: 0.4192 - val_accuracy: 0.8305\n",
      "Epoch 6/10\n",
      "102/102 [==============================] - 16s 160ms/step - loss: 0.2093 - accuracy: 0.9132 - val_loss: 0.4318 - val_accuracy: 0.8332\n",
      "Epoch 7/10\n",
      "102/102 [==============================] - 16s 155ms/step - loss: 0.2000 - accuracy: 0.9174 - val_loss: 0.4394 - val_accuracy: 0.8308\n",
      "Epoch 8/10\n",
      "102/102 [==============================] - 16s 154ms/step - loss: 0.1908 - accuracy: 0.9223 - val_loss: 0.4588 - val_accuracy: 0.8313\n",
      "Epoch 9/10\n",
      "102/102 [==============================] - 16s 154ms/step - loss: 0.1818 - accuracy: 0.9266 - val_loss: 0.4700 - val_accuracy: 0.8314\n",
      "Epoch 10/10\n",
      "102/102 [==============================] - 16s 155ms/step - loss: 0.1724 - accuracy: 0.9313 - val_loss: 0.4833 - val_accuracy: 0.8298\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(train_data, epochs=10, validation_data=valid_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history.history.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 261,
       "width": 391
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'], label='train loss')\n",
    "plt.plot(history.history['val_loss'], label='validation loss')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 261,
       "width": 391
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['accuracy'], label='train accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='validation accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### As we can see, the model training accuracy keep on increasing w.r.t the epochs but the validation accuracy reached a peak value and start decreasing again. And similiar pattern is observed for the loss function but in the opposite sense. This is related to the overfitting problem.\n",
    "#### There are two ways we can solve this problem.\n",
    "<li>Use drop layers in between the layer</li>\n",
    "<li>Stop training the model before the loss value starts to increase</li>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_model1():\n",
    "    model = tf.keras.Sequential()\n",
    "    model.add(hub_layer)\n",
    "    model.add(tf.keras.layers.Dense(16, activation='relu'))\n",
    "    model.add(tf.keras.layers.Dropout(0.3))\n",
    "    model.add(tf.keras.layers.Dense(16, activation='relu'))\n",
    "    model.add(tf.keras.layers.Dropout(0.3))\n",
    "    model.add(tf.keras.layers.Dense(1, activation='sigmoid'))\n",
    "    model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=0.001),\n",
    "              loss=tf.keras.losses.BinaryCrossentropy(),\n",
    "              metrics=['accuracy'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "model1 = get_model1()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Here we are using earlystop callback to stop training when the loss on the validation data is reached the minimum value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "102/102 [==============================] - 16s 150ms/step - loss: 0.6160 - accuracy: 0.6723 - val_loss: 0.4544 - val_accuracy: 0.8220\n",
      "Epoch 2/20\n",
      "102/102 [==============================] - 15s 148ms/step - loss: 0.4045 - accuracy: 0.8358 - val_loss: 0.3786 - val_accuracy: 0.8326\n",
      "Epoch 3/20\n",
      "102/102 [==============================] - 15s 148ms/step - loss: 0.3282 - accuracy: 0.8695 - val_loss: 0.3807 - val_accuracy: 0.8307\n",
      "Epoch 4/20\n",
      "102/102 [==============================] - 15s 149ms/step - loss: 0.2939 - accuracy: 0.8838 - val_loss: 0.3965 - val_accuracy: 0.8287\n",
      "Epoch 5/20\n",
      "102/102 [==============================] - 15s 148ms/step - loss: 0.2688 - accuracy: 0.8932 - val_loss: 0.4123 - val_accuracy: 0.8282\n",
      "Epoch 5: early stopping\n"
     ]
    }
   ],
   "source": [
    "es = EarlyStopping(monitor='val_loss', mode='auto', verbose=1, patience=3)\n",
    "history = model1.fit(train_data, epochs=20, validation_data=valid_data, callbacks=[es])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 261,
       "width": 391
      },
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['accuracy'], label='train accuracy')\n",
    "plt.plot(history.history['val_accuracy'], label='validation accuracy')\n",
    "plt.xlabel('Epoch')\n",
    "plt.xticks(history.epoch)\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(history.history['val_accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### After the earlystop callback is triggered, the model is not trained anymore and maximum accuracy is achieved at 1st epoch training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13/13 [==============================] - 0s 22ms/step - loss: 0.4250 - accuracy: 0.8215\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.42498770356178284, 0.8215110301971436]"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model1.evaluate(test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13/13 [==============================] - 0s 23ms/step - loss: 0.5228 - accuracy: 0.8149\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.5227763056755066, 0.81488037109375]"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(valid_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "machine_learning",
   "language": "python",
   "name": "machine_learning"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "25d94ae763beb282c8936bac0f3ff970d91daf0b8f2bdf99ab1c21bc0960c6c6"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}