homework/Day_042_HW.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [作業重點]\n",
    "目前你應該已經要很清楚資料集中，資料的型態是什麼樣子囉！包含特徵 (features) 與標籤 (labels)。因此要記得未來不管什麼專案，必須要把資料清理成相同的格式，才能送進模型訓練。\n",
    "今天的作業開始踏入決策樹這個非常重要的模型，請務必確保你理解模型中每個超參數的意思，並試著調整看看，對最終預測結果的影響為何"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 作業\n",
    "\n",
    "1. 試著調整 DecisionTreeClassifier(...) 中的參數，並觀察是否會改變結果？\n",
    "2. 改用其他資料集 (boston, wine)，並與回歸模型的結果進行比較"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 建立模型四步驟\n",
    "\n",
    "在 Scikit-learn 中，建立一個機器學習的模型其實非常簡單，流程大略是以下四個步驟\n",
    "\n",
    "1. 讀進資料，並檢查資料的 shape (有多少 samples (rows), 多少 features (columns)，label 的型態是什麼？)\n",
    "    - 讀取資料的方法：\n",
    "        - **使用 pandas 讀取 .csv 檔：**pd.read_csv\n",
    "        - **使用 numpy 讀取 .txt 檔：**np.loadtxt \n",
    "        - **使用 Scikit-learn 內建的資料集：**sklearn.datasets.load_xxx\n",
    "    - **檢查資料數量：**data.shape (data should be np.array or dataframe)\n",
    "2. 將資料切為訓練 (train) / 測試 (test)\n",
    "    - train_test_split(data)\n",
    "3. 建立模型，將資料 fit 進模型開始訓練\n",
    "    - clf = DecisionTreeClassifier()\n",
    "    - clf.fit(x_train, y_train)\n",
    "4. 將測試資料 (features) 放進訓練好的模型中，得到 prediction，與測試資料的 label (y_test) 做評估\n",
    "    - clf.predict(x_test)\n",
    "    - accuracy_score(y_test, y_pred)\n",
    "    - f1_score(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets, metrics\n",
    "\n",
    "# 如果是分類問題，請使用 DecisionTreeClassifier，若為回歸問題，請使用 DecisionTreeRegressor\n",
    "from sklearn.tree import DecisionTreeClassifier, DecisionTreeRegressor\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 讀取波士頓房價資料集\n",
    "boston = datasets.load_boston()\n",
    "\n",
    "# 切分訓練集/測試集\n",
    "x_train, x_test, y_train, y_test = train_test_split(boston.data, boston.target, test_size=0.25, random_state=4)\n",
    "\n",
    "# 建立模型\n",
    "reg = DecisionTreeRegressor()\n",
    "\n",
    "# 訓練模型\n",
    "reg.fit(x_train, y_train)\n",
    "\n",
    "# 預測測試集\n",
    "y_pred = reg.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MAE:  25.523779527559054\n"
     ]
    }
   ],
   "source": [
    "mse = metrics.mean_squared_error(y_test, y_pred)\n",
    "print(\"MAE: \", mse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['CRIM' 'ZN' 'INDUS' 'CHAS' 'NOX' 'RM' 'AGE' 'DIS' 'RAD' 'TAX' 'PTRATIO'\n",
      " 'B' 'LSTAT']\n"
     ]
    }
   ],
   "source": [
    "print(boston.feature_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature importance:  [7.37215324e-02 1.87787040e-03 1.61548333e-02 5.43024196e-05\n",
      " 2.90488853e-02 5.42020934e-01 1.36261608e-02 5.87021305e-02\n",
      " 4.74328133e-04 9.87726768e-03 1.97277195e-02 1.15299599e-02\n",
      " 2.23184076e-01]\n"
     ]
    }
   ],
   "source": [
    "print(\"Feature importance: \", reg.feature_importances_)"
   ]
  }
 ],
 "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
}