benchmark/TP2AEC-AVALIACAO.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Critérios de Avaliação"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Acurácia\n",
    "\n",
    "Recall:\n",
    "- O recall da classe 1 e da classe 0 não pode ser muito diferente.\n",
    "- Se a recall da classe 1 supera a da classe 0, existe um overfit na classe 1. Vice-versa\n",
    "- Ver \"somas\" das classes do lado esquerdo e do lado direito. Se a diferenca for muito grande, significa que o modelo classifica melhor uma das classes.\n",
    "- Deve existir um maior erro a prever classe 0 do que classe 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle as pkl\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score, confusion_matrix, classification_report\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from keras.models import load_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test = pd.read_csv('./../models/datasets/df_test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_test_norm = pd.read_csv('./../models/datasets/test_normalize.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Para os seguintes modelos foi usado o dataset tratado, mas sem ser normalizado"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_cm(y_true, y_pred, figsize=(6,6)):\n",
    "    cm = confusion_matrix(y_true, y_pred, labels=np.unique(y_true))\n",
    "    cm_sum = np.sum(cm, axis=1, keepdims=True)\n",
    "    cm_perc = cm / cm_sum.astype(float) * 100\n",
    "    annot = np.empty_like(cm).astype(str)\n",
    "    nrows, ncols = cm.shape\n",
    "    for i in range(nrows):\n",
    "        for j in range(ncols):\n",
    "            c = cm[i, j]\n",
    "            p = cm_perc[i, j]\n",
    "            if i == j:\n",
    "                s = cm_sum[i]\n",
    "                annot[i, j] = '%.1f%%\\n%d/%d' % (p, c, s)\n",
    "            elif c == 0:\n",
    "                annot[i, j] = ''\n",
    "            else:\n",
    "                annot[i, j] = '%.1f%%\\n%d' % (p, c)\n",
    "    cm = pd.DataFrame(cm, index=np.unique(y_true), columns=np.unique(y_true))\n",
    "    cm.index.name = 'Actual'\n",
    "    cm.columns.name = 'Predicted'\n",
    "    fig, ax = plt.subplots(figsize=figsize)\n",
    "    ax.xaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n",
    "    ax.yaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n",
    "    sns.heatmap(cm, cmap= \"YlGnBu\", annot=annot, fmt='', ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def showModel(df_test, model):\n",
    "    X_test = df_test.drop(\"target\", axis = 1)\n",
    "    y_test = df_test.target\n",
    "    y_pred = model.predict(X_test)\n",
    "    acc = accuracy_score(y_test, y_pred)\n",
    "    print('Acurácia:\\n', acc)\n",
    "    print('Report dos parametros:\\n')\n",
    "    print (classification_report(y_test, y_pred))\n",
    "    print('Matriz de confusão:\\n')\n",
    "    plot_cm(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo K-Nearest-Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "KNN = pkl.load(open('models/modeloKNN.pkl','rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acurácia:\n",
      " 0.8282868525896414\n",
      "Report dos parametros:\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.88      0.89      0.89     11360\n",
      "           1       0.66      0.63      0.64      3700\n",
      "\n",
      "    accuracy                           0.83     15060\n",
      "   macro avg       0.77      0.76      0.76     15060\n",
      "weighted avg       0.83      0.83      0.83     15060\n",
      "\n",
      "Matriz de confusão:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-d4a32cf3cb7c>:22: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.xaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n",
      "<ipython-input-4-d4a32cf3cb7c>:23: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.yaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "showModel(data_test,KNN)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "RandomForest = pkl.load(open('models/modeloRandomForest.pkl','rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acurácia:\n",
      " 0.8088977423638778\n",
      "Report dos parametros:\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.92      0.82      0.87     11360\n",
      "           1       0.58      0.79      0.67      3700\n",
      "\n",
      "    accuracy                           0.81     15060\n",
      "   macro avg       0.75      0.80      0.77     15060\n",
      "weighted avg       0.84      0.81      0.82     15060\n",
      "\n",
      "Matriz de confusão:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-d4a32cf3cb7c>:22: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.xaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n",
      "<ipython-input-4-d4a32cf3cb7c>:23: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.yaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "showModel(data_test,RandomForest)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo Naive-Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "NaiveBayes= pkl.load(open(\"models/modeloNaiveBayes.pkl\",'rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acurácia:\n",
      " 0.701261620185923\n",
      "Report dos parametros:\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.65      0.77     11360\n",
      "           1       0.44      0.86      0.59      3700\n",
      "\n",
      "    accuracy                           0.70     15060\n",
      "   macro avg       0.69      0.76      0.68     15060\n",
      "weighted avg       0.81      0.70      0.72     15060\n",
      "\n",
      "Matriz de confusão:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-d4a32cf3cb7c>:22: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.xaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n",
      "<ipython-input-4-d4a32cf3cb7c>:23: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.yaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "showModel(data_test,NaiveBayes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo KMeans"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Devido aos péssimos resultados obtidos por este modelo, decidimos não guardar tendo em conta que é certo que não vai ser o melhor modelo a usar usado e eventualmente a ser \"deployed\" para um mercado de trabalho."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Para os seguintes modelos foi usado o dataset normalizado"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo Regressão Logística"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "RegLog= pkl.load(open('models/modeloRegLog.pkl','rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-d4a32cf3cb7c>:22: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.xaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acurácia:\n",
      " 0.803519256308101\n",
      "Report dos parametros:\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.92      0.81      0.86     11360\n",
      "           1       0.57      0.79      0.66      3700\n",
      "\n",
      "    accuracy                           0.80     15060\n",
      "   macro avg       0.75      0.80      0.76     15060\n",
      "weighted avg       0.84      0.80      0.81     15060\n",
      "\n",
      "Matriz de confusão:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-d4a32cf3cb7c>:23: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.yaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "showModel(data_test_norm, RegLog)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo SVM (Support Vector Machine)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "SVM = pkl.load(open('models/modeloSVM.pkl','rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acurácia:\n",
      " 0.7933598937583002\n",
      "Report dos parametros:\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.94      0.78      0.85     11360\n",
      "           1       0.55      0.84      0.67      3700\n",
      "\n",
      "    accuracy                           0.79     15060\n",
      "   macro avg       0.75      0.81      0.76     15060\n",
      "weighted avg       0.84      0.79      0.81     15060\n",
      "\n",
      "Matriz de confusão:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-d4a32cf3cb7c>:22: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.xaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n",
      "<ipython-input-4-d4a32cf3cb7c>:23: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.yaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "showModel(data_test_norm, SVM)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Modelo Artificial Neural Networks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "471/471 [==============================] - 0s 409us/step - loss: 0.4002 - accuracy: 0.7909\n",
      "Accuracy: \n",
      " 0.7909030318260193\n",
      "\n",
      "Classification Report\n",
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.78      0.85     11360\n",
      "           1       0.55      0.82      0.66      3700\n",
      "\n",
      "    accuracy                           0.79     15060\n",
      "   macro avg       0.74      0.80      0.75     15060\n",
      "weighted avg       0.84      0.79      0.80     15060\n",
      "\n",
      "\n",
      " Matriz de Confusão:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-4-d4a32cf3cb7c>:22: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.xaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n",
      "<ipython-input-4-d4a32cf3cb7c>:23: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
      "  ax.yaxis.set_ticklabels(['<=50K/Y', '>50K/Y'])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_test = data_test_norm.drop(\"target\", axis = 1)\n",
    "y_test = data_test_norm.target\n",
    "## To load model \n",
    "model = load_model('models/modelANN.h5')\n",
    "## To load model IMPORT ===== from keras.models import load_model\n",
    "y_pred = model.predict(X_test)\n",
    "y_pred = (y_pred > 0.5)\n",
    "score = model.evaluate(X_test, y_test,verbose=1)\n",
    "print('Accuracy: \\n',score[1])\n",
    "print (\"\\nClassification Report\\n\")\n",
    "print (classification_report(y_test, y_pred))\n",
    "print(\"\\n Matriz de Confusão:\\n\")\n",
    "plot_cm(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusões"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Com estes resultados conclui-se que se tivermos a analisar só em termos de acurácia, o modelo ideal seria o K-Nearest-Neighbors pois é o modelo que tem a melhor acurácia (~83%).\n",
    "\n",
    "No entanto, o nosso grupo decidiu ter em conta o recall (como pode ser visto na secção de Critérios de Avaliação) para decidir qual o melhor modelo a ser utilizado. Com isto, conclui-mos que o melhor modelo será o modelo de Regressão Logística, pois tem uma acurácia alta e a matriz de confusão deste é a que apresenta melhores resultados."
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}