{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "_This notebook contains code and comments from Section 4.4 of the book [Ensemble Methods for Machine Learning](https://www.manning.com/books/ensemble-methods-for-machine-learning). Please see the book for additional details on this topic. This notebook and code are released under the [MIT license](https://github.com/gkunapuli/ensemble-methods-notebooks/blob/master/LICENSE)._\n", "\n", "---\n", "\n", "## 4.4 Case Study: Handwriting Digit Classification\n", "In this case study, we will use scikit-learn’s digits data set to illustrate the effectiveness of AdaBoost. The data set consists of 1797 scanned images of handwritten digits from 0 to 9. Each digit is associated with a unique label, which makes this a 10-class classification problem. There are roughly 180 digits per class.\n", "\n", "The digits themselves are represented as 16 x 16 normalized greyscale bitmap images, which when flattened results in a 64-dimensional vector for each handwritten digit. Thus, the training set is of size 1797 (examples) x 64 features. " ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(1797, 64)" ] }, "execution_count": 1, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.datasets import load_digits\n", "X, y = load_digits(return_X_y=True)\n", "X.shape" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "import numpy as np" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can visualize this data set as a snapshot to get an idea of what they look like." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 360x360 with 1 Axes>" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(nrows=1, ncols=1, figsize=(5, 5))\n", "\n", "n_img_per_row = 8\n", "img = np.zeros((10 * n_img_per_row, 10 * n_img_per_row))\n", "for i in range(n_img_per_row):\n", " ix = 10 * i + 1\n", " for j in range(n_img_per_row):\n", " iy = 10 * j + 1\n", " img[ix:ix + 8, iy:iy + 8] = X[i * n_img_per_row + j].reshape((8, 8))\n", "\n", "ax.imshow(img, cmap=plt.cm.binary)\n", "ax.set_xticks([])\n", "ax.set_yticks([])\n", "ax.axis('off')\n", "# ax.set_title('A selection from the 64-dimensional digits dataset')\n", "\n", "# plt.savefig('./figures/CH04_F13_Kunapuli.png', format='png', dpi=300, bbox_inches='tight', pad_inches=0);\n", "# plt.savefig('./figures/CH04_F13_Kunapuli.pdf', format='pdf', dpi=300, bbox_inches='tight', pad_inches=0);" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "---\n", "### 4.4.1\tDimensionality Reduction with t-SNE\n", "\n", "While AdaBoost can effectively handle the dimensionality of the digits data set (64 features), we will (rather aggressively) look to reduce the dimensionality to 2. The main reason for this is to be able to visualize the data as well as the models learned by AdaBoost.\n", "\n", "We’ll use a nonlinear dimensionality reduction technique known as [t-distributed stochastic neighbor](https://lvdmaaten.github.io/tsne/) embedding or t-SNE. t-SNE is a highly effective pre-processing technique for the digits data set and extracts an effective embedding in a two-dimensional space.\n", "\n", "In ``scikit-learn``, the [``manifold.TSNE``](https://scikit-learn.org/stable/modules/generated/sklearn.manifold.TSNE.html) package implements t-SNE." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "C:\\Python310\\lib\\site-packages\\sklearn\\manifold\\_t_sne.py:805: FutureWarning: The default learning rate in TSNE will change from 200.0 to 'auto' in 1.2.\n", " warnings.warn(\n", "C:\\Python310\\lib\\site-packages\\sklearn\\manifold\\_t_sne.py:991: FutureWarning: The PCA initialization in TSNE will change to have the standard deviation of PC1 equal to 1e-4 in 1.2. This will ensure better convergence.\n", " warnings.warn(\n" ] } ], "source": [ "from sklearn.manifold import TSNE\n", "Xemb = TSNE(n_components=2, init='pca').fit_transform(X)" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 432x432 with 1 Axes>" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(nrows=1, ncols=1, figsize=(6, 6))\n", "xMin, xMax = np.min(Xemb, axis=0), np.max(Xemb, axis=0)\n", "Xemb = (Xemb - xMin) / (xMax - xMin)\n", "\n", "for i in range(Xemb.shape[0]):\n", " if np.random.ranf() < 0.8:\n", " # Skip some data points so that the image isn't super cluttered\n", " continue\n", " plt.text(Xemb[i, 0], Xemb[i, 1], str(y[i]), color=plt.cm.tab10(y[i] / 10.),\n", " fontdict={'size': 18})\n", "\n", "ax.set_xticks([])\n", "ax.set_yticks([])\n", "ax.axis('off')\n", "fig.tight_layout()\n", "\n", "# plt.savefig('./figures/CH04_F14_Kunapuli.png', format='png', dpi=300, bbox_inches='tight', pad_inches=0)\n", "# plt.savefig('./figures/CH04_F14_Kunapuli.pdf', format='pdf', dpi=300, bbox_inches='tight', pad_inches=0)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It is important to hold aside a part of the training data for evaluation and to quantify the predictive performance of our models on future data. We split the lower-dimensional data Xemb and the labels into training and test sets.\n", "\n", "Observe the use of ``stratify=y``, to ensure that the ratios of the different digits in train and test sets are identical." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "from sklearn.model_selection import train_test_split\n", "Xtrn, Xtst, ytrn, ytst = train_test_split(Xemb, y, test_size=0.2, stratify=y, random_state=13)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "---\n", "### 4.4.2\tBoosting\n", "\n", "We will now train an AdaBoost model for this digit classification task. Recall from our earlier discussion that AdaBoost requires us to first decide the type of base estimator. We continue to use decision stumps." ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "from sklearn.tree import DecisionTreeClassifier\n", "from sklearn.ensemble import AdaBoostClassifier\n", "\n", "stump = DecisionTreeClassifier(max_depth=2)\n", "ensemble = AdaBoostClassifier(algorithm='SAMME', base_estimator=stump)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To identify the best combination of ``learning_rate`` and ``n_estimators`` for the ``AdaBoostClassifier``, we will employ a combination of k-fold cross validation and [grid search](https://en.wikipedia.org/wiki/Hyperparameter_optimization#Grid_search). \n", "\n", "The basic idea is to consider different combinations of ``learning_rate`` and ``n_estimators`` and evaluate what their performance would be like via cross validation. First, we select various parameter values we want to explore." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "parameters_to_search = {'n_estimators': [200, 300, 400, 500],\n", " 'learning_rate': [0.6, 0.8, 1.0]}" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Next, we make a scoring function to evaluate the performance of each parameter combination. For this task, we use the balanced accuracy score, which is essentially just the accuracy score weighted by each class. " ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "from sklearn.metrics import balanced_accuracy_score, make_scorer\n", "scorer = make_scorer(balanced_accuracy_score, greater_is_better=True)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now, we set up and run the grid search to identify the best parameter combination with the [``GridSearchCV``](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html) class. The parameter ``cv=5`` specifies 5-fold cross validation and ``n_jobs=-1`` specifies that the job should use all available cores for parallel processing. \n", "\n", "The final parameter in ``GridSearchCV`` is set to ``refit=True``. This tells GridSearchCV to train a final model using all the available training data using the best parameter combination it has identified." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "text/html": [ "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=5,\n", " estimator=AdaBoostClassifier(algorithm='SAMME',\n", " base_estimator=DecisionTreeClassifier(max_depth=2)),\n", " n_jobs=-1,\n", " param_grid={'learning_rate': [0.6, 0.8, 1.0],\n", " 'n_estimators': [200, 300, 400, 500]},\n", " scoring=make_scorer(balanced_accuracy_score))</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=5,\n", " estimator=AdaBoostClassifier(algorithm='SAMME',\n", " base_estimator=DecisionTreeClassifier(max_depth=2)),\n", " n_jobs=-1,\n", " param_grid={'learning_rate': [0.6, 0.8, 1.0],\n", " 'n_estimators': [200, 300, 400, 500]},\n", " scoring=make_scorer(balanced_accuracy_score))</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: AdaBoostClassifier</label><div class=\"sk-toggleable__content\"><pre>AdaBoostClassifier(algorithm='SAMME',\n", " base_estimator=DecisionTreeClassifier(max_depth=2))</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">base_estimator: DecisionTreeClassifier</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeClassifier(max_depth=2)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">DecisionTreeClassifier</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeClassifier(max_depth=2)</pre></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div>" ], "text/plain": [ "GridSearchCV(cv=5,\n", " estimator=AdaBoostClassifier(algorithm='SAMME',\n", " base_estimator=DecisionTreeClassifier(max_depth=2)),\n", " n_jobs=-1,\n", " param_grid={'learning_rate': [0.6, 0.8, 1.0],\n", " 'n_estimators': [200, 300, 400, 500]},\n", " scoring=make_scorer(balanced_accuracy_score))" ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.model_selection import GridSearchCV\n", "search = GridSearchCV(ensemble, param_grid=parameters_to_search,\n", " scoring=scorer, cv=5, n_jobs=-1, refit=True)\n", "search.fit(Xtrn, ytrn)" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "The best parameter settings are {'learning_rate': 1.0, 'n_estimators': 300}, with score = 0.9394760080277322.\n" ] } ], "source": [ "best_combo = search.cv_results_['params'][search.best_index_]\n", "best_score = search.best_score_\n", "print('The best parameter settings are {0}, with score = {1}.'.format(best_combo, best_score))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The best model is available in ``search.best_estimator_`` and can be used for making predictions on the test data." ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "ypred = search.best_estimator_.predict(Xtst)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "How well did this model do? We can first look at the classification report." ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Classification report:\n", " precision recall f1-score support\n", "\n", " 0 1.00 0.97 0.99 36\n", " 1 1.00 1.00 1.00 37\n", " 2 1.00 0.97 0.99 35\n", " 3 1.00 1.00 1.00 37\n", " 4 0.97 1.00 0.99 36\n", " 5 0.72 1.00 0.84 36\n", " 6 1.00 1.00 1.00 36\n", " 7 1.00 1.00 1.00 36\n", " 8 0.95 1.00 0.97 35\n", " 9 1.00 0.58 0.74 36\n", "\n", " accuracy 0.95 360\n", " macro avg 0.96 0.95 0.95 360\n", "weighted avg 0.96 0.95 0.95 360\n", "\n", "\n" ] } ], "source": [ "from sklearn.metrics import classification_report\n", "print('Classification report:\\n{0}\\n'.format(classification_report(ytst, ypred)))" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Confusion matrix: \n", " [[35 0 0 0 1 0 0 0 0 0]\n", " [ 0 37 0 0 0 0 0 0 0 0]\n", " [ 0 0 34 0 0 0 0 0 1 0]\n", " [ 0 0 0 37 0 0 0 0 0 0]\n", " [ 0 0 0 0 36 0 0 0 0 0]\n", " [ 0 0 0 0 0 36 0 0 0 0]\n", " [ 0 0 0 0 0 0 36 0 0 0]\n", " [ 0 0 0 0 0 0 0 36 0 0]\n", " [ 0 0 0 0 0 0 0 0 35 0]\n", " [ 0 0 0 0 0 14 0 0 1 21]]\n" ] } ], "source": [ "from sklearn.metrics import confusion_matrix\n", "print(\"Confusion matrix: \\n {0}\".format(confusion_matrix(ytst, ypred)))" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "<Figure size 432x432 with 1 Axes>" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Visualize the decision boundary\n", "xMin, xMax = Xemb[:, 0].min(), Xemb[:, 0].max() + 0.05\n", "yMin, yMax = Xemb[:, 1].min(), Xemb[:, 1].max() + 0.05\n", "xMesh, yMesh = np.meshgrid(np.arange(xMin, xMax, 0.05),\n", " np.arange(yMin, yMax, 0.05))\n", "\n", "fig, ax = plt.subplots(nrows=1, ncols=1, figsize=(6, 6))\n", "zMesh = search.best_estimator_.predict(np.c_[xMesh.ravel(), yMesh.ravel()])\n", "zMesh = zMesh.reshape(xMesh.shape) * 1.0\n", "# boundary = ax.contourf(xMesh, yMesh, zMesh, np.array([-1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9]),\n", "# cmap=plt.cm.tab10, alpha=0.15)\n", "boundary = ax.contour(xMesh, yMesh, zMesh, np.array([-1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9]), cmap=plt.cm.tab10, alpha=0.5)\n", " \n", "for i in range(X.shape[0]):\n", " if np.random.ranf() < 0.8:\n", " # Skip some data points so that the image isn't super cluttered\n", " continue\n", " plt.text(Xemb[i, 0], Xemb[i, 1], str(y[i]), color=plt.cm.tab10(y[i] / 10.),\n", " fontdict={'size': 18})\n", "\n", "ax.axis('off')\n", "# fig.colorbar(boundary)\n", "\n", "fig.tight_layout()\n", "# plt.savefig('./figures/CH04_F15_Kunapuli.png', format='png', dpi=200, bbox_inches='tight', pad_inches=0)\n", "# plt.savefig('./figures/CH04_F15_Kunapuli.pdf', format='pdf', dpi=200, bbox_inches='tight', pad_inches=0)" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "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": "3.10.5" } }, "nbformat": 4, "nbformat_minor": 2 }