{ "cells": [ { "cell_type": "markdown", "id": "c42b15c3-8172-4dc9-afeb-67adc7df5587", "metadata": {}, "source": [ "# Compare Classifier Package Tutorial\n", "A short tutorial on comparing performances of multiple classifiers with Voting or Stacking ensemble methods. **Please note this package only works within the world of scikit-learn.**" ] }, { "cell_type": "markdown", "id": "c565464b-60bd-41ae-9a51-806d694d19f6", "metadata": {}, "source": [ "## The Prerequisites" ] }, { "cell_type": "markdown", "id": "e9f6d2a1-a68c-4650-b6fa-d7cc70e4106b", "metadata": {}, "source": [ "### Step 1: Prepare the Data\n", "\n", "We are using a downloaded version of the [UCI Wine Quality dataset](https://archive.ics.uci.edu/dataset/186/wine+quality), named 'example_dataset.csv', stored in the project root directory. It contains physicochemical features as numerical values of wines and a target variable, column `quality` representing wine quality score. Our task is to train models to predict wine quality scores, integer values from 3 to 9 (9 = highest quality; 3 = lowest quality).\n", "\n", "> _**Just In Case You're Wondering:**_ We changed an original dataset column, `color` containing text indicating a wine being 'red' or 'white' into a binary column now named `is_red` with 1 indicating red wine, and 0 indicating white wine. This ensures all columns are numeric and facilitates the upcoming training process for our models.\n", "\n", "We will also need to split the dataset into `X_train`, `y_train`, `X_test`, `y_test` based on features and target variable columns, with a test dataset consisting 20% of all data. Here are some [explanations on data splitting](https://www.geeksforgeeks.org/how-to-split-the-dataset-with-scikit-learns-train_test_split-function/), if you need them. \n", "\n", "Now, let's take a peek at `X_train`, our training data without the target variables:" ] }, { "cell_type": "code", "execution_count": 2, "id": "b5809064-5419-495e-b492-c7b5cbcaef56", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
is_redfixed_acidityvolatile_aciditycitric_acidresidual_sugarchloridesfree_sulfur_dioxidetotal_sulfur_dioxidedensitypHsulphatesalcohol
125615.60.540.041.70.0495.013.00.994203.720.5811.4
396306.60.240.3011.30.02611.077.00.993813.130.5512.8
312106.50.320.231.20.05439.0208.00.992723.180.469.9
53119.60.600.502.30.07928.071.00.999703.500.579.7
54617.90.350.211.90.07346.0102.00.996403.270.589.5
.......................................
345406.50.290.311.70.03524.079.00.990533.270.6911.4
153806.10.410.1410.40.03718.0119.00.996003.380.4510.0
8714.70.600.172.30.05817.0106.00.993203.850.6012.9
147506.30.360.304.80.04914.085.00.993203.280.3910.6
419207.10.210.352.50.04041.0186.00.991283.320.5612.5
\n", "

4089 rows × 12 columns

\n", "
" ], "text/plain": [ " is_red fixed_acidity volatile_acidity citric_acid residual_sugar \\\n", "1256 1 5.6 0.54 0.04 1.7 \n", "3963 0 6.6 0.24 0.30 11.3 \n", "3121 0 6.5 0.32 0.23 1.2 \n", "531 1 9.6 0.60 0.50 2.3 \n", "546 1 7.9 0.35 0.21 1.9 \n", "... ... ... ... ... ... \n", "3454 0 6.5 0.29 0.31 1.7 \n", "1538 0 6.1 0.41 0.14 10.4 \n", "87 1 4.7 0.60 0.17 2.3 \n", "1475 0 6.3 0.36 0.30 4.8 \n", "4192 0 7.1 0.21 0.35 2.5 \n", "\n", " chlorides free_sulfur_dioxide total_sulfur_dioxide density pH \\\n", "1256 0.049 5.0 13.0 0.99420 3.72 \n", "3963 0.026 11.0 77.0 0.99381 3.13 \n", "3121 0.054 39.0 208.0 0.99272 3.18 \n", "531 0.079 28.0 71.0 0.99970 3.50 \n", "546 0.073 46.0 102.0 0.99640 3.27 \n", "... ... ... ... ... ... \n", "3454 0.035 24.0 79.0 0.99053 3.27 \n", "1538 0.037 18.0 119.0 0.99600 3.38 \n", "87 0.058 17.0 106.0 0.99320 3.85 \n", "1475 0.049 14.0 85.0 0.99320 3.28 \n", "4192 0.040 41.0 186.0 0.99128 3.32 \n", "\n", " sulphates alcohol \n", "1256 0.58 11.4 \n", "3963 0.55 12.8 \n", "3121 0.46 9.9 \n", "531 0.57 9.7 \n", "546 0.58 9.5 \n", "... ... ... \n", "3454 0.69 11.4 \n", "1538 0.45 10.0 \n", "87 0.60 12.9 \n", "1475 0.39 10.6 \n", "4192 0.56 12.5 \n", "\n", "[4089 rows x 12 columns]" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import pandas as pd\n", "from sklearn.model_selection import train_test_split\n", " \n", "example_dataset = pd.read_csv('../example_dataset.csv')\n", "\n", "# convert the `color` column to a binary variable: `is_red`: red = 1, white = 0, and drop the original `color` column\n", "example_dataset['is_red'] = example_dataset['color'].apply(lambda x: 1 if x == 'red' else 0)\n", "example_dataset = example_dataset.drop(['color'], axis=1)\n", "\n", "# move the `color` column to the beginning\n", "last_col = example_dataset.pop(example_dataset.columns[-1])\n", "example_dataset.insert(0, last_col.name, last_col)\n", "\n", "train_df, test_df = train_test_split(example_dataset, test_size=0.2)\n", "\n", "X_train, X_test, y_train, y_test = (\n", " train_df.drop(columns='quality'), test_df.drop(columns='quality'),\n", " train_df['quality'], test_df['quality']\n", ")\n", "\n", "X_train" ] }, { "cell_type": "markdown", "id": "7a7b5da8-deac-41ce-93c4-9d1396d5a8cc", "metadata": {}, "source": [ "### Step 2: Prepare Your Models\n", "\n", "Then, we will need to build your scikit-learn machine learning models and perform [hyperparameter tuning](https://scikit-learn.org/1.5/modules/grid_search.html) to make sure they are doing the best they can. We will assume we have 5 final candidates as below. Our package functions accept the models in a list of (name, model) tuple, so we will need to convert them into something like this:\n", "\n", "> _**Naming the Models:**_ The names you gave the models will appear in function outputs as labels. For example, for the first model in `multi_ind` below, if you want to see something more elaborate and descriptive of the model, you could write the tuple as `('Logistic Regression (C=92)', logreg)`. For simplicity's sake, we are giving our models very concise names in this tutorial." ] }, { "cell_type": "code", "execution_count": 3, "id": "d1425a44-e040-4921-beb7-693b854025a7", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[('logreg', LogisticRegression(C=92, multi_class='multinomial')),\n", " ('gb', GradientBoostingClassifier(learning_rate=0.001, n_estimators=1000)),\n", " ('svm', SVC(max_iter=2000)),\n", " ('rf', RandomForestClassifier(n_estimators=10)),\n", " ('knn5', KNeighborsClassifier())]" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.svm import SVC\n", "from sklearn.linear_model import LogisticRegression\n", "from sklearn.ensemble import GradientBoostingClassifier\n", "from sklearn.neighbors import KNeighborsClassifier\n", "from sklearn.ensemble import RandomForestClassifier\n", "\n", "logreg = LogisticRegression(multi_class='multinomial', solver='lbfgs', C=92)\n", "gb = GradientBoostingClassifier(learning_rate=0.001, n_estimators=1000)\n", "svm = SVC(kernel='rbf', decision_function_shape='ovr', max_iter=2000)\n", "rf = RandomForestClassifier(n_estimators=10)\n", "knn5 = KNeighborsClassifier(n_neighbors=5)\n", "\n", "multi_ind = [\n", " ('logreg', logreg),\n", " ('gb', gb),\n", " ('svm', svm),\n", " ('rf', rf),\n", " ('knn5', knn5)\n", "]\n", "\n", "multi_ind" ] }, { "cell_type": "markdown", "id": "029bc434-47e9-4875-9be4-8123445c281b", "metadata": {}, "source": [ "> _**A Note on Pipelines:**_ For simplicity's sake, all the models used in this tutorial are individual sklearn classifiers. However, sklearn pipelines are also accepted by our functions, so below is valid input models as well." ] }, { "cell_type": "code", "execution_count": 4, "id": "b39a36a7-717e-495c-b5ab-97379d45d271", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[('pipe_svm',\n", " Pipeline(steps=[('robustscaler', RobustScaler()), ('svc', SVC(max_iter=2000))])),\n", " ('pipe_rf',\n", " Pipeline(steps=[('robustscaler', RobustScaler()),\n", " ('randomforestclassifier',\n", " RandomForestClassifier(n_estimators=10))])),\n", " ('pipe_knn5',\n", " Pipeline(steps=[('robustscaler', RobustScaler()),\n", " ('kneighborsclassifier', KNeighborsClassifier())])),\n", " ('pipe_gb',\n", " Pipeline(steps=[('robustscaler', RobustScaler()),\n", " ('gradientboostingclassifier',\n", " GradientBoostingClassifier(learning_rate=0.001,\n", " n_estimators=1000))])),\n", " ('pipe_logreg',\n", " Pipeline(steps=[('robustscaler', RobustScaler()),\n", " ('logisticregression',\n", " LogisticRegression(C=92, multi_class='multinomial'))]))]" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from sklearn.preprocessing import RobustScaler\n", "from sklearn.pipeline import make_pipeline\n", "\n", "pipe_svm = make_pipeline(RobustScaler(), svm)\n", "pipe_rf = make_pipeline(RobustScaler(), rf)\n", "pipe_knn5 = make_pipeline(RobustScaler(), knn5)\n", "pipe_gb = make_pipeline(RobustScaler(), gb)\n", "pipe_logreg = make_pipeline(RobustScaler(), logreg)\n", "\n", "multi_pipe = [\n", " ('pipe_svm', pipe_svm),\n", " ('pipe_rf', pipe_rf),\n", " ('pipe_knn5', pipe_knn5),\n", " ('pipe_gb', pipe_gb),\n", " ('pipe_logreg', pipe_logreg)\n", "]\n", "\n", "multi_pipe" ] }, { "cell_type": "markdown", "id": "61404832-5cb3-4c37-b8a3-9488f0e137c5", "metadata": {}, "source": [ "## Let's Get Started!" ] }, { "cell_type": "markdown", "id": "a28a3b41-2f6d-44b7-b800-a4617c9ab8fd", "metadata": {}, "source": [ "We recommend following all steps below for informed decision making when choosing a model for your classification task." ] }, { "cell_type": "markdown", "id": "1f673bf1-6d86-42a1-b6c4-a5f146601eba", "metadata": {}, "source": [ "### Step 1: Compare Confusion Matrices for Candidate Models\n", "\n", "A [confusion matrix](https://www.geeksforgeeks.org/confusion-matrix-machine-learning/) is used to evaluate the performance of a classification model by comparing its predicted values against the actual values, providing a detailed breakdown of correct and incorrect predictions, allowing for a deeper analysis of the model's strengths and weaknesses beyond just overall accuracy. We can take the models we've built in `multi_ind` from the **Prerequisites - Step 3: Prepare Your Models** section and look at how each is performing by looking at their confusion matrices side-by-side to compare them." ] }, { "cell_type": "code", "execution_count": 7, "id": "4baaa137-ec5b-44c6-ac75-174e7c14b968", "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# We have not dealt with warning messages due to the time limit, so we will surpress them for now.\n", "import warnings\n", "warnings.filterwarnings(\"ignore\")\n", "\n", "from compare_classifiers.confusion_matrices import confusion_matrices\n", "\n", "confusion_matrices(multi_ind, X_train, X_test, y_train, y_test);" ] }, { "cell_type": "markdown", "id": "703e9852-112e-42b4-b09d-d77d2128bc57", "metadata": {}, "source": [ "### Step 2: Compare Fit Time and F1 Scores for Candidate Models\n", "\n", "[F1 score](https://www.v7labs.com/blog/f1-score-guide) is a machine learning metric that measures a model's accuracy by combining its precision and recall by measuring the harmonic mean and is commonly used as an evaluation metric in binary and multi-class classification. Our package uses a macro-averaged F1 score (or macro F1 score). It is computed by taking the arithmetic mean (aka unweighted mean) of all the per-class F1 scores. This method treats all classes equally regardless of their support values.\n", "\n", "Let's take a look at the fit times and f1 scores using 5-fold cross validation." ] }, { "cell_type": "code", "execution_count": 23, "id": "502217a1-bea8-41f8-931a-fa4c8adce71c", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
modelfit_timetest_f1_scoretrain_f1_score
0logreg0.0792830.1665000.156686
1gb39.2275010.2033100.425625
2svm0.4918380.0951700.089807
3rf0.0696190.2560100.968338
4knn50.0031490.1956650.332405
\n", "
" ], "text/plain": [ " model fit_time test_f1_score train_f1_score\n", "0 logreg 0.079283 0.166500 0.156686\n", "1 gb 39.227501 0.203310 0.425625\n", "2 svm 0.491838 0.095170 0.089807\n", "3 rf 0.069619 0.256010 0.968338\n", "4 knn5 0.003149 0.195665 0.332405" ] }, "execution_count": 23, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from compare_classifiers.compare_f1 import compare_f1\n", "\n", "compare_f1(multi_ind, X_train, y_train)" ] }, { "cell_type": "markdown", "id": "4b0c0af4-5a05-4f9a-ab97-fb2971e9de7c", "metadata": {}, "source": [ "_(Please excuse the low test score! Due to limited time of the project, we are using random hyperparamters and was not expecting impressing results.)_ " ] }, { "cell_type": "markdown", "id": "ae5471bf-8b6d-4316-9dcf-3616334c1959", "metadata": {}, "source": [ "### Step 3: Compare Fit Time and F1 Scores Through Voting and Stacking Ensembles\n", "\n", "The goal of sklearn [ensemble methods](https://medium.com/@abhishekjainindore24/different-types-of-ensemble-techniques-bagging-boosting-stacking-voting-blending-b04355a03c93) is to combine the predictions of several base estimators built with a given learning algorithm in order to improve generalizability / robustness over a single estimator. There are different types of combination methods to build ensembles and each one yields slightly different results. Our package has two ensembles available from sklearn: [Voting](https://scikit-learn.org/1.5/modules/generated/sklearn.ensemble.VotingClassifier.html) and [Stacking](https://scikit-learn.org/dev/modules/generated/sklearn.ensemble.StackingClassifier.html) with default settings. \n", "\n", "Assume we've chosen the following models: `logreg`, `svm` and `knn5` as our finalists as they show the least amounts of overfitting. We can then take a look at their performances with both ensemble methods, again using 5-fold cross validation." ] }, { "cell_type": "code", "execution_count": 24, "id": "97b17e3b-f55c-4fe2-9dd2-54833776ec92", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
methodfit_timetest_f1_scoretrain_f1_score
0voting0.5651260.1431950.173578
1stacking3.3983040.1822370.180828
\n", "
" ], "text/plain": [ " method fit_time test_f1_score train_f1_score\n", "0 voting 0.565126 0.143195 0.173578\n", "1 stacking 3.398304 0.182237 0.180828" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from compare_classifiers.ensemble_compare_f1 import ensemble_compare_f1\n", "\n", "finalists = [\n", " ('logreg', logreg),\n", " ('svm', svm),\n", " ('knn5', knn5)\n", "]\n", "\n", "ensemble_compare_f1(finalists, X_train, y_train)" ] }, { "cell_type": "markdown", "id": "034ba086-85ea-4f6a-8b5e-27a2b3d6d2d5", "metadata": {}, "source": [ "> _**Results:**_ Each run might be different and you may now find yourself with a winning ensemble method here, or that both Voting and Stacking perform similarly." ] }, { "cell_type": "markdown", "id": "fe1aacf0-a45a-43a0-aceb-7a8e7a40553d", "metadata": {}, "source": [ "### Step 4: Compare Ensemble Results With Individual Classifier Results" ] }, { "cell_type": "markdown", "id": "df6ae6d3-bbf8-4cc8-9cd6-eda6abb67dac", "metadata": {}, "source": [ "This step does not require running any package functions, but we believe it is an essential step in the comparing model performances. Since the Voting f1 score does not show a significant increase from some of our classifiers' individual f1 scores (looking at results from Step 2 above), we might want to assess whether we should choose to go forward with an ensemble model or without.\n", "\n", "**Now, if you have decided to go with one of the existing individual classifiers you built** (because sometimes individual models can perform almost equally well as an ensemble and people may prefer using a single model because it is easier to re-train and manipulate)**, then you have already ended up with the choice of a best model!**\n", "\n", "However, if you have decided to use an ensemble to predict instead, then in the next section, we will demonstrate how to use the Voting ensemble method to predict the target classes. " ] }, { "cell_type": "markdown", "id": "3bd23a69-ccfa-47b4-8444-c00e3de14fc2", "metadata": {}, "source": [ "## (Optional) Predict Using Ensemble" ] }, { "cell_type": "markdown", "id": "8d567b14-547c-481d-9e51-d293e185b549", "metadata": {}, "source": [ "### Predict Unseen Data With Chosen Ensemble Method\n", "\n", "Now it's time to put our winner ensemble method to test and have it predict some test or unseen data. We will predict on `X_test` using Voting ensemble for demonstration purpose.\n", "\n", "> _**Note:**_ If you would like to use Stacking as the ensemble method, you would pass in 'stacking' instead of 'voting' instead." ] }, { "cell_type": "code", "execution_count": 29, "id": "cadd1020-ff8b-443c-801c-ad01963dc209", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([6, 6, 6, ..., 6, 6, 6], shape=(1023,))" ] }, "execution_count": 29, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from compare_classifiers.ensemble_predict import ensemble_predict\n", "\n", "ensemble_predict(finalists, X_train, y_train, 'voting', X_test)" ] }, { "cell_type": "markdown", "id": "272c718f-e695-4805-a752-d20a428e72bc", "metadata": {}, "source": [ "These are the wine quality scores we get for `X_test` with our Voting ensemble classifier from the three finalist models we defined in Step 3, and here are the unique quality scores predicted!" ] }, { "cell_type": "code", "execution_count": 30, "id": "f284a120-1fc1-4396-acbf-b39b7149a14b", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([4, 5, 6, 7])" ] }, "execution_count": 30, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import numpy as np\n", "\n", "predictions = ensemble_predict(finalists, X_train, y_train, 'voting', X_test)\n", "\n", "np.unique(predictions)" ] } ], "metadata": { "kernelspec": { "display_name": "Python [conda env:compare_classifiers]", "language": "python", "name": "conda-env-compare_classifiers-py" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.11.11" } }, "nbformat": 4, "nbformat_minor": 5 }