22_early_stopping_gradient_boosting.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`Gradient Boosting` is an ensemble technique where several weak regression trees are combined to yield a powerful model <br>\n",
    "`Early stopping` enables us to find the least number of iterations sufficient to build that generalises well on unseen data"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In `Early Stopping` we specify a `validation_fraction` - denoting a fraction of the whole dataset. When each additional stage of regression tree is added, the validation set is used to score the model, and continue it till the score do not improve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd \n",
    "import numpy as np \n",
    "import matplotlib.pyplot as plt \n",
    "%matplotlib inline \n",
    "\n",
    "from sklearn import ensemble \n",
    "from sklearn import datasets \n",
    "from sklearn.model_selection import train_test_split\n",
    "import time "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_list = [\n",
    "    datasets.load_iris(return_X_y=True),\n",
    "    datasets.make_classification(n_samples=800, random_state=0),\n",
    "    datasets.make_hastie_10_2(n_samples=2000, random_state=0)\n",
    "]\n",
    "names = [\"Iris Data\", \"Classification Data\",\"Hastie Data\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_gb = []\n",
    "score_gb = []\n",
    "time_gb = []\n",
    "n_gbes = []\n",
    "score_gbes = []\n",
    "time_gbes = []\n",
    "\n",
    "n_estimators = 200"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "for X, y in data_list:\n",
    "    \n",
    "    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)\n",
    "    \n",
    "    #If the score don't improve by at 0.01 for the last 10 stages, stop additional stages \n",
    "    gbes = ensemble.GradientBoostingClassifier(\n",
    "        n_estimators=n_estimators,\n",
    "        validation_fraction=0.2,\n",
    "        n_iter_no_change=5,\n",
    "        tol=0.01,\n",
    "        random_state=0\n",
    "    )\n",
    "    gb = ensemble.GradientBoostingClassifier(n_estimators=n_estimators, random_state=0)\n",
    "    \n",
    "    start = time.time()\n",
    "    gb.fit(X_train, y_train)\n",
    "    time_gb.append(time.time() - start)\n",
    "\n",
    "    start = time.time()\n",
    "    gbes.fit(X_train, y_train)\n",
    "    time_gbes.append(time.time() - start)\n",
    "\n",
    "    score_gb.append(gb.score(X_test, y_test))\n",
    "    score_gbes.append(gbes.score(X_test, y_test))\n",
    "\n",
    "    n_gb.append(gb.n_estimators)\n",
    "    n_gbes.append(gbes.n_estimators)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, comparing score with and without early stopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 648x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "bar_width =0.2 \n",
    "n = len(data_list)\n",
    "index = np.arange(0, n * bar_width, bar_width) * 2.5\n",
    "index = index[0:n]\n",
    "plt.figure(figsize=(9,5))\n",
    "\n",
    "bar_1 = plt.bar(index, score_gb, bar_width, label=\"Without Early Stopping\",color=\"blue\")\n",
    "bar_2 = plt.bar(index + bar_width, score_gbes, bar_width, label=\"With Early Stopping\",color=\"green\")\n",
    "\n",
    "plt.xticks(index + bar_width, names)\n",
    "plt.yticks(np.arange(0,1.3,0.1))\n",
    "\n",
    "def autolabel(rects, n_estimators):\n",
    "    \"\"\"\n",
    "    Attach a text label above each bar displaying n_estimators of each model\n",
    "    \"\"\"\n",
    "    for i, rect in enumerate(rects):\n",
    "        plt.text(\n",
    "            rect.get_x() + rect.get_width() / 2.0,\n",
    "            1.05 * rect.get_height(),\n",
    "            \"n_est=%d\" % n_estimators[i],\n",
    "            ha=\"center\",\n",
    "            va=\"bottom\",\n",
    "        )\n",
    "\n",
    "autolabel(bar_1, n_gb)\n",
    "autolabel(bar_2, n_gbes)\n",
    "\n",
    "plt.ylim([0, 1.3])\n",
    "plt.legend(loc=\"best\")\n",
    "plt.grid(True)\n",
    "\n",
    "plt.xlabel(\"Datasets\")\n",
    "plt.ylabel(\"Test score\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Comparing model fitting time with and without early stopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 648x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(9, 5))\n",
    "\n",
    "bar1 = plt.bar(index, time_gb, bar_width, label=\"Without early stopping\", color=\"blue\")\n",
    "bar2 = plt.bar(index + bar_width, time_gbes, bar_width, label=\"With early stopping\", color=\"green\")\n",
    "\n",
    "max_y = np.amax(np.maximum(time_gb, time_gbes))\n",
    "\n",
    "plt.xticks(index + bar_width, names)\n",
    "plt.yticks(np.linspace(0, 1.3 * max_y, 13))\n",
    "\n",
    "autolabel(bar1, n_gb)\n",
    "autolabel(bar2, n_gbes)\n",
    "\n",
    "plt.ylim([0, 1.3 * max_y])\n",
    "plt.legend(loc=\"best\")\n",
    "plt.grid(True)\n",
    "\n",
    "plt.xlabel(\"Datasets\")\n",
    "plt.ylabel(\"Fit Time\")\n",
    "\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.9.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}