6.Revaluation_and_Deployment.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "# 6. Revaluation and Deployment"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 6.1 Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\candy\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\deprecation.py:144: FutureWarning: The sklearn.metrics.classification module is  deprecated in version 0.22 and will be removed in version 0.24. The corresponding classes / functions should instead be imported from sklearn.metrics. Anything that cannot be imported from sklearn.metrics is now part of the private API.\n",
      "  warnings.warn(message, FutureWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\candy\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\compat\\v2_compat.py:96: disable_resource_variables (from tensorflow.python.ops.variable_scope) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "non-resource variables are not supported in the long term\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import pickle\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "from Helpers_NN import conf_matrix\n",
    "from Classification import Classification\n",
    "\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras.models import load_model\n",
    "import tensorflow.compat.v1 as tf\n",
    "tf.disable_v2_behavior() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "sns.set_context('poster')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\candy\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\keras\\initializers.py:119: calling RandomUniform.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n",
      "WARNING:tensorflow:From C:\\Users\\candy\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\ops\\resource_variable_ops.py:1666: calling BaseResourceVariable.__init__ (from tensorflow.python.ops.resource_variable_ops) with constraint is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "If using Keras pass *_constraint arguments to layers.\n",
      "WARNING:tensorflow:From C:\\Users\\candy\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\ops\\init_ops.py:97: calling GlorotUniform.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n",
      "WARNING:tensorflow:From C:\\Users\\candy\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\ops\\init_ops.py:97: calling Zeros.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n"
     ]
    }
   ],
   "source": [
    "stacking_model = pickle.load(open('Models/Stacking.pkl','rb'))\n",
    "neural_net_model =  load_model('Models/Neural_Network.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "x_test = pd.read_csv('Data/4.x_test_data.csv')\n",
    "y_test = pd.read_csv('Data/4.y_test_data.csv')\n",
    "x_test_array = np.load('Data/6.x_test.npy')\n",
    "y_test_ohe = pd.read_csv('Data/6.y_test_ohe.csv')\n",
    "y_test_nn = pd.read_csv('Data/6.y_test_nn_df.csv')\n",
    "y_test_predicted_array = np.load('Data/6.y_test_predicted_array.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [],
   "source": [
    "neural_net_model.save('Models/Neural_Network.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 6.2 Comparing Stacking Model with Neural Network Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 6.2.1 Test Accuracy for Stacking Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "Collapsed": "false"
   },
   "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>Model Name</th>\n",
       "      <th>Train Accuracy</th>\n",
       "      <th>Validation Accuracy</th>\n",
       "      <th>Accuracy Difference</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Stacking</td>\n",
       "      <td>0.55467</td>\n",
       "      <td>0.507709</td>\n",
       "      <td>0.046961</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Model Name  Train Accuracy  Validation Accuracy  Accuracy Difference\n",
       "0   Stacking         0.55467             0.507709             0.046961"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>precision</td>\n",
       "      <td>0.626344</td>\n",
       "      <td>0.441667</td>\n",
       "      <td>0.417098</td>\n",
       "      <td>0.460000</td>\n",
       "      <td>0.600551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>recall</td>\n",
       "      <td>0.619681</td>\n",
       "      <td>0.447887</td>\n",
       "      <td>0.428191</td>\n",
       "      <td>0.473008</td>\n",
       "      <td>0.566234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>f1-score</td>\n",
       "      <td>0.622995</td>\n",
       "      <td>0.444755</td>\n",
       "      <td>0.422572</td>\n",
       "      <td>0.466413</td>\n",
       "      <td>0.582888</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  1         2         3         4         5\n",
       "precision  0.626344  0.441667  0.417098  0.460000  0.600551\n",
       "recall     0.619681  0.447887  0.428191  0.473008  0.566234\n",
       "f1-score   0.622995  0.444755  0.422572  0.466413  0.582888"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stacking_model.get_test_scores(x_test,y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 6.2.2 Test Accuracy for Neural Network Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "I tested the neural network model using the test data and achieved an accuracy of <strong>0.5710</strong> which is better than the stacking model accuracy of <strong>0.5077</strong>, by <strong>over 5%</strong>. \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.9816686870695618, 0.5624668]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neural_net_model.evaluate(x_test_array,y_test_ohe)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 6.2.3 Confusion Matrix for Stacking Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "I wanted to look at the confusion matrix, as this gives a better idea of how the model is performing over all 5 classes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "stacking_model.test_conf_matrix()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 6.2.4 Confusion Matrix for Neural Network Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "Collapsed": "false"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_cm = confusion_matrix(y_test_nn['score'], y_test_predicted_array)\n",
    "conf_matrix(test_cm)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "The error is more contained with in adjacent scores with the neural network model. Almost zero confusion between extreme scores 1 and 5, and minimal confusion with scores 2 and 4. Although a score of 3 can be harder to predict, there is definitely an improvement from the Stacking model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "## 6.3 Deployment"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 6.3.1 Streamlit Application"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "After seeing the improvements from the Stacking model, I was more confident about deploying the model for actionable use."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "I planned on future improvements being the addition of the neural network model and then creating an application for the model, so as a next step I decided to make a working application to test out new reviews using streamlit."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "### 6.3.2 Future Development"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "Collapsed": "false"
   },
   "source": [
    "- Create a webscraper spider for twitter, reddit, etc for further testing"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}