Predição Inadimplência Cartão de Crédito.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predição Inadimplência Cartão de Crédito"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Descrição dos dados\n",
    "\n",
    "* *X1:* Quanto foi dado de crédito em dólar (individual e família) (montante de crédito concedido)\n",
    "* *X2:* Gênero -> 1 = Masculino, 2 = Feminino\n",
    "* *X3:* Educação -> 1 = Pós-Graduação, 2 = Universidade, 3 = Ensino Médio, 4 = Outros\n",
    "* *X4:* Estado Civil -> 1 = Casado, 2 = Solteiro, 3 = Outros\n",
    "* *X5:* Idade\n",
    "* *X6 - X11:* Histórico de Pagamentos -> X6 = Setembro à X11 = Abril (PAY_...)\n",
    "    * -1: Pagou devidamente\n",
    "    *  1: Pagou com atraso de um mês\n",
    "    *  2: Pagou com atraso de dois meses\n",
    "    *  3: Pagou com atraso de três meses\n",
    "    ...\n",
    "    \n",
    "* *X12 - X17:* Quantia de declaração de conta -> X12 = Setembro à X17 = Abril (BILL_AMT_...)\n",
    "* *X18 -> X23:* Montante do pagamento anterior -> X18 = Setembro à X23 = Abril (PAY_AMT_...)\n",
    "* *X24:* Se a pessoa pagou, variável de interesse -> 1 = Sim, 2 = Não"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "df = pd.read_excel('default of credit card clients.xls')\n",
    "labels = df.iloc[0].values\n",
    "df = df.iloc[1:]\n",
    "\n",
    "df = pd.get_dummies(df, columns=['X2', 'X3', 'X4'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc72ede1898>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Heatmap with original data\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "sns.set()\n",
    "df = df.astype(float)\n",
    "\n",
    "plt.figure(figsize=(10, 10))\n",
    "sns.heatmap(df.corr(), square=True, cmap='RdYlGn')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop records with minimum data\n",
    "df.drop(inplace=True, columns=['X3_0', 'X4_0', 'X3_5', 'X3_6', 'X15', 'X16', 'X17'])\n",
    "\n",
    "# Substitute discrete values - Mode\n",
    "for i in range(6, 12):\n",
    "    x_idx = f'X{str(i)}'\n",
    "    \n",
    "    np_array = df[x_idx].values\n",
    "    mode = pd.Series(np_array[np_array != 0]).mode()[0]\n",
    "    \n",
    "    df[x_idx].replace(-2, np.nan, inplace=True)\n",
    "    df[x_idx].fillna(mode, inplace=True)\n",
    "\n",
    "# # Substitute discrete values - Median\n",
    "# for i in range(6, 12):\n",
    "#     x_idx = f'X{str(i)}'\n",
    "    \n",
    "#     df[x_idx].replace(0, np.nan, inplace=True)\n",
    "#     df[x_idx].fillna(df[x_idx].median(skipna=True), inplace=True)\n",
    "\n",
    "# Substitute real values\n",
    "for i in range(12, 24):\n",
    "    if i not in [15, 16, 17]:\n",
    "        x_idx = f'X{str(i)}'\n",
    "        sum_value = pd.to_numeric(df[x_idx], errors='coerce').sum()\n",
    "        amount = len(df[x_idx].nonzero()[0])\n",
    "        mean = sum_value/amount\n",
    "\n",
    "        df[x_idx].replace(0, mean, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc740cadcf8>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Heatmap with sliced data\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "sns.set()\n",
    "df = df.astype(float)\n",
    "\n",
    "plt.figure(figsize=(10, 10))\n",
    "sns.heatmap(df.corr(), square=True, cmap='RdYlGn')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split training data and test data\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "target = df['Y'].values\n",
    "data = df.drop('Y', axis=1).values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    data, target, test_size=0.35, random_state=42, stratify=target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "        PAID       0.79      0.97      0.87      8177\n",
      "    NOT PAID       0.48      0.09      0.16      2323\n",
      "\n",
      "   micro avg       0.78      0.78      0.78     10500\n",
      "   macro avg       0.63      0.53      0.51     10500\n",
      "weighted avg       0.72      0.78      0.71     10500\n",
      "\n",
      "0.776952380952381\n"
     ]
    }
   ],
   "source": [
    "# Using KNN mathematical model to calculate performance.\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "knn = KNeighborsClassifier(n_neighbors=10)\n",
    "knn.fit(X_train, y_train)\n",
    "\n",
    "y_pred = knn.predict(X_test)\n",
    "\n",
    "print(classification_report(y_test, y_pred, target_names=['PAID', 'NOT PAID']))\n",
    "print(knn.score(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Barplot of Y    \n",
    "plt.figure(figsize=(10,8))\n",
    "barlist = plt.bar(df.Y.unique(),df.Y.value_counts())\n",
    "barlist[0].set_color('r')\n",
    "plt.xticks(df.Y.unique(),('Unpaid','Paid'))\n",
    "plt.ylabel('count')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Barplot of Payment History\n",
    "plt.figure(figsize=(10,10))\n",
    "barlist = plt.bar(df.X6.unique(),df.X6.value_counts())\n",
    "barlist[0].set_color('r')\n",
    "plt.xticks(df.X6.unique(),str(df.X6.unique()).replace('[', '').replace(']', '').split('. '))\n",
    "plt.ylabel('Count')\n",
    "plt.xlabel('Payment months delay')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sklearn.manifold import TSNE\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "X = df.drop(['Y'],axis=1)\n",
    "X = df.head(1000)\n",
    "\n",
    "y = df['Y'].head(1000)\n",
    "y = y.values\n",
    "\n",
    "tsne = TSNE(n_components=2, random_state=0)\n",
    "\n",
    "X_2d = tsne.fit_transform(X.values)\n",
    "target_ids = range(len((df.Y.unique())))\n",
    "\n",
    "colors = 'r', 'g', 'b', 'c', 'm', 'y', 'k', 'w', 'orange', 'purple'\n",
    "for i, c, label in zip(target_ids, colors, ('NÃO PAGOU', 'PAGOU')):\n",
    "    plt.scatter(X_2d[y == i, 0], X_2d[y == i, 1], c=c, label=label)\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Resultados\n",
    "\n",
    "### Resultados dos modelos escolhidos\n",
    "|      Modelo      | Decision Tree | SGD  | Logistic Regression | Naive Bayes | Neural Networks | Gradient Tree Boosting | KNN (baseline) |\n",
    "| ---------------- |:-------------:|:----:|:-------------------:|:-----------:|:---------------:|:------------:|:------------:|\n",
    "| Precisão (Pagou) |       83%     |  78% |         82%         |     84%     |       71%       |      83%     |      79%     |\n",
    "| Acurácia         |       72%     |  72% |         81%         |     78%     |       21%       |      81%     |      77%     |"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Resultados de referência\n",
    "![Tabela de resultados do artigo](images/paper_results_table.png)"
   ]
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}