stacking-classifier-ensemble-for-great-results.ipynb

{"cells":[{"source":"<a href=\"https://www.kaggle.com/code/marcinrutecki/stacking-classifier-ensemble-for-great-results?scriptVersionId=119837176\" target=\"_blank\"><img align=\"left\" alt=\"Kaggle\" title=\"Open in Kaggle\" src=\"https://kaggle.com/static/images/open-in-kaggle.svg\"></a>","metadata":{},"cell_type":"markdown","outputs":[],"execution_count":0},{"cell_type":"markdown","id":"e8b5546a","metadata":{"papermill":{"duration":0.019781,"end_time":"2023-02-21T09:32:15.733512","exception":false,"start_time":"2023-02-21T09:32:15.713731","status":"completed"},"tags":[]},"source":["Telco Churn - EDA & Model (Voting + boosting)<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:200%;\n","            text-align:center;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">♞ Stacking classifier - ensemble model for great results ♞</div>"]},{"cell_type":"code","execution_count":1,"id":"c6ecc383","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:15.770443Z","iopub.status.busy":"2023-02-21T09:32:15.769932Z","iopub.status.idle":"2023-02-21T09:32:15.780725Z","shell.execute_reply":"2023-02-21T09:32:15.779544Z"},"papermill":{"duration":0.032342,"end_time":"2023-02-21T09:32:15.783375","exception":false,"start_time":"2023-02-21T09:32:15.751033","status":"completed"},"tags":[]},"outputs":[],"source":["from IPython.display import Image\n","#Image by https://www.pexels.com/pl-pl/zdjecie/sztuka-kobiety-granie-muzyka-7097455/"]},{"attachments":{"fd470f41-6ec6-4f81-bf43-b452e8bd7f7c.jpg":{"image/jpeg":"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"}},"cell_type":"markdown","id":"5d3b438b","metadata":{"papermill":{"duration":0.01719,"end_time":"2023-02-21T09:32:15.818295","exception":false,"start_time":"2023-02-21T09:32:15.801105","status":"completed"},"tags":[]},"source":["![ensemble.jpg](attachment:fd470f41-6ec6-4f81-bf43-b452e8bd7f7c.jpg)"]},{"cell_type":"code","execution_count":2,"id":"36bdf52a","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:15.855885Z","iopub.status.busy":"2023-02-21T09:32:15.855452Z","iopub.status.idle":"2023-02-21T09:32:16.7823Z","shell.execute_reply":"2023-02-21T09:32:16.780781Z"},"papermill":{"duration":0.948886,"end_time":"2023-02-21T09:32:16.78529","exception":false,"start_time":"2023-02-21T09:32:15.836404","status":"completed"},"tags":[]},"outputs":[],"source":["from matplotlib import pyplot as plt\n","import seaborn as sns"]},{"cell_type":"markdown","id":"6aa3dab6","metadata":{"papermill":{"duration":0.019186,"end_time":"2023-02-21T09:32:16.822843","exception":false,"start_time":"2023-02-21T09:32:16.803657","status":"completed"},"tags":[]},"source":["<a id=\"1.2\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:170%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\"><span style='color:#CDA63A'>|</span> Table of Contents</div>"]},{"cell_type":"markdown","id":"309d4b57","metadata":{"papermill":{"duration":0.017985,"end_time":"2023-02-21T09:32:16.859828","exception":false,"start_time":"2023-02-21T09:32:16.841843","status":"completed"},"tags":[]},"source":["<a id=\"toc\"></a>\n","- [1. Unity is strength - introduction](#1)\n","    - [1.1 Voting vs Stacking](#1.1)\n","    - [1.2 Stacking vs bagging and boosting](#1.2)\n","    - [1.3 Framework](#1.3)\n","    - [1.4 Tuning hyperparameters in ensemble learning](#1.4)\n","- [2. Set-up](#2)\n","    - [2.1 Import Libraries](#2.1)\n","    - [2.2 Import Data](#2.2)\n","    - [2.3 Data Set Characteristics](#2.3)\n","    - [2.4 Dataset Attributes](#2.4)\n","- [3. Data preprocessing](#3)\n","    - [3.1 Dealin with missing values in TotalCharges](#3.1)\n","    - [3.2 Dealing with duplicate values](#3.2)\n","    - [3.3 Creating numerical and categorical lists](#3.3)\n","- [4. Feature Engineering](#4)\n","    - [4.1 Train test split - stratified splitting](#4.1)\n","    - [4.2 Feature scaling](#4.2)\n","    - [4.3 One hot Encoding](#4.3)\n","        - [4.3.1 X_train encoding](#4.3.1)\n","        - [4.3.2 X_test encoding](#4.3.2)\n","- [5. Model building](#5)\n","    - [5.1 Level-0 with 28 Models](#5.1)\n","    - [5.2 Level-1 Model](#5.2)\n","    - [5.3 Create a stacking classifier](#5.3)\n","    - [5.4 AUC-ROC Curve](#5.4)\n","    - [5.5 Precision-Recall Curve (PR curve)](#5.5)\n","    - [5.6 This is just the beginning](#5.6)\n","- [6. My other notebooks](#6)\n","- [7. References](#7)"]},{"cell_type":"markdown","id":"b4d01001","metadata":{"papermill":{"duration":0.017728,"end_time":"2023-02-21T09:32:16.896414","exception":false,"start_time":"2023-02-21T09:32:16.878686","status":"completed"},"tags":[]},"source":["<a id=\"1\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:170%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">1 <span style='color:#CDA63A'>|</span> Unity is strength - introduction</div>"]},{"cell_type":"markdown","id":"78f48ebc","metadata":{"papermill":{"duration":0.017443,"end_time":"2023-02-21T09:32:16.932181","exception":false,"start_time":"2023-02-21T09:32:16.914738","status":"completed"},"tags":[]},"source":["A stacking classifier is an ensemble method where the output from multiple classifiers is passed as an input to a meta-classifier for the task of the final classification. The individual classification models are trained based on the complete training set, then the meta-classifier is fitted based on the outputs (meta-features) of the individual classification models."]},{"cell_type":"markdown","id":"d6bd066c","metadata":{"papermill":{"duration":0.019036,"end_time":"2023-02-21T09:32:16.968983","exception":false,"start_time":"2023-02-21T09:32:16.949947","status":"completed"},"tags":[]},"source":["\n","<div style=\" background-color:#3b3745;\n","            text-align:left;\n","            margin: auto;\n","            padding: 13px 13px; \n","            border-radius: 8px; \n","            color: white\">\n","<ul>\n","<b>Ensemble learning</b> is a machine learning technique that combines several base models in order to produce an optimal predictive model.\n","\n","<b>Stacking</b> has been successfully used in several machine learning competitions at Kaggle. It is definitely a must know technique in machine learning.\n","</ul>\n","</div>"]},{"attachments":{"5a8fdb1a-94b9-49a2-b3ad-81217759ec61.png":{"image/png":"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"}},"cell_type":"markdown","id":"f2c1b64d","metadata":{"papermill":{"duration":0.017893,"end_time":"2023-02-21T09:32:17.004712","exception":false,"start_time":"2023-02-21T09:32:16.986819","status":"completed"},"tags":[]},"source":["![stack.png](attachment:5a8fdb1a-94b9-49a2-b3ad-81217759ec61.png)"]},{"cell_type":"markdown","id":"1d750c07","metadata":{"papermill":{"duration":0.018211,"end_time":"2023-02-21T09:32:17.041349","exception":false,"start_time":"2023-02-21T09:32:17.023138","status":"completed"},"tags":[]},"source":["<div style=\"background-color:#CDA63A;\n","            color:white;\n","            text-align:left;\n","            width:70%;\n","            padding: 13px 13px; \n","            border-radius: 8px; \n","            margin: auto\">\n","\n","<b>In essence:</b> stacking algorithm uses multiple models to make baseline predictions, and using these models, trains a Meta-Classifier for better performance. Stacking allows us to use the strength of each individual estimator by taking advantage of their output as input for the final estimator. Therefore stacking can make predictions that have better performance than any single model in the ensemble.\n","</div>"]},{"cell_type":"markdown","id":"7f56769e","metadata":{"papermill":{"duration":0.017813,"end_time":"2023-02-21T09:32:17.077265","exception":false,"start_time":"2023-02-21T09:32:17.059452","status":"completed"},"tags":[]},"source":["<a id=\"1.1\"></a>\n","## <b>1.1 <span style='color:#E1B12D'>Voting vs Stacking</span></b>"]},{"cell_type":"markdown","id":"d2fdc552","metadata":{"papermill":{"duration":0.017758,"end_time":"2023-02-21T09:32:17.112922","exception":false,"start_time":"2023-02-21T09:32:17.095164","status":"completed"},"tags":[]},"source":["The fundamental difference between voting and stacking is how the final aggregation works. Voting aglomerates results of classifier's decisions by taking the class which appears in most cases or by weights of the predictions a whereas stacking performs the final aggregation by using a blender/meta classifier."]},{"cell_type":"markdown","id":"540c706a","metadata":{"papermill":{"duration":0.018031,"end_time":"2023-02-21T09:32:17.149044","exception":false,"start_time":"2023-02-21T09:32:17.131013","status":"completed"},"tags":[]},"source":["<div class=\"alert alert-block alert-info\" style=\"background-color:#3b3745;\n","                                                 color:white;\n","                                                 border-color:black;\n","                                                 border-radius:5px;\n","                                                 margin: auto;\n","                                                 text-align: left;\">\n","<b>Voting </b> - the algorithm has some constant method that gives answers.\n","    \n","<b>Stacking </b> - the algorithm takes advantage of the predictions as the new representation of the problem and creates another abstraction layer to learn how to predict the correct label having k votes.\n","</div>"]},{"cell_type":"markdown","id":"92197e32","metadata":{"papermill":{"duration":0.017859,"end_time":"2023-02-21T09:32:17.184994","exception":false,"start_time":"2023-02-21T09:32:17.167135","status":"completed"},"tags":[]},"source":["Check this notebook for **<mark style=\"background-color:#CDA63A;color:white;border-radius:5px;opacity:1.0\">MUCH more information</mark>** **<mark style=\"background-color:#3b3745;color:white;border-radius:5px;opacity:1.0\">about voting classifier</mark>**:\n","\n","* [Voting Classifier for Better Results](https://www.kaggle.com/code/marcinrutecki/voting-classifier-for-better-results)"]},{"cell_type":"markdown","id":"7c0d9020","metadata":{"papermill":{"duration":0.017491,"end_time":"2023-02-21T09:32:17.220865","exception":false,"start_time":"2023-02-21T09:32:17.203374","status":"completed"},"tags":[]},"source":["<a id=\"1.2\"></a>\n","## <b>1.2 <span style='color:#E1B12D'>Stacking vs bagging and boosting</span></b>"]},{"cell_type":"markdown","id":"20038c77","metadata":{"papermill":{"duration":0.017655,"end_time":"2023-02-21T09:32:17.256377","exception":false,"start_time":"2023-02-21T09:32:17.238722","status":"completed"},"tags":[]},"source":["1. **Bagging** (stands for Bootstrap Aggregating): we use bagging for combining weak learners (base models) of high variance. Bagging aims to produce a model with lower variance than the individual weak models. Bagging takes advantage of  Bootstrapping technique - sampling different sets of data from a given training set by using replacement. After bootstrapping the training dataset, we train the model on all the different sets and aggregate the result. Unlike bagging, in stacking, the models are typically different (e.g. not all decision trees) and fit on the same dataset (instead of samples of the training dataset).\n","\n","\n","2. **Boosting**: in boosting the learners are trained sequentially. The algorithm learns models sequentially in a very adaptive way (a base model depends on the previous ones) and combines them following a deterministic strategy. Unlike boosting, in stacking, a single model is used to learn how to best combine the predictions from the contributing models (e.g. instead of correcting the predictions of prior models).\n","\n","Bagging trains models in parallel, boosting trains the models sequentially. Stacking creates a new meta-model."]},{"cell_type":"markdown","id":"c6c2d737","metadata":{"papermill":{"duration":0.017371,"end_time":"2023-02-21T09:32:17.291359","exception":false,"start_time":"2023-02-21T09:32:17.273988","status":"completed"},"tags":[]},"source":["<a id=\"1.3\"></a>\n","## <b>1.3 <span style='color:#E1B12D'>Framework</span></b>"]},{"cell_type":"markdown","id":"0d499bce","metadata":{"papermill":{"duration":0.017291,"end_time":"2023-02-21T09:32:17.326282","exception":false,"start_time":"2023-02-21T09:32:17.308991","status":"completed"},"tags":[]},"source":["The framework of a stacked ensemble consists of two or more base models (level-0 models) and a higher level meta model (level-1 model).\n","\n","<div class=\"alert alert-block alert-info\" style=\"background-color:#3b3745;\n","                                                 color:white;\n","                                                 border-color:black;\n","                                                 border-radius:5px;\n","                                                 margin: auto;\n","                                                 text-align: left;\">\n","<b>Base-Models </b> (Level-0 Models): models that fit the training data and predict out-of-sample data.\n","    \n","<b>Meta-Model </b> (Level-1 Model): final estimator that fits on the prediction from base-models and learns how to best combine the predictions (the default classifier is a LogisticRegression.).\n","</div>"]},{"cell_type":"markdown","id":"13e03ff4","metadata":{"papermill":{"duration":0.017384,"end_time":"2023-02-21T09:32:17.361248","exception":false,"start_time":"2023-02-21T09:32:17.343864","status":"completed"},"tags":[]},"source":["<div style=\"background-color:#CDA63A;\n","            color:white;\n","            text-align:left;\n","            width:70%;\n","            padding: 13px 13px; \n","            border-radius: 8px; \n","            margin: auto\">\n","\n","<b>In essence:</b> stacking algorithm uses multiple models to make baseline predictions, and using these models, trains a Meta-Classifier for better performance. Stacking allows us to use the strength of each individual estimator by taking advantage of their output as input for the final estimator. Therefore stacking can make predictions that have better performance than any single model in the ensemble.\n","\n","</div>"]},{"cell_type":"markdown","id":"53a25292","metadata":{"papermill":{"duration":0.017329,"end_time":"2023-02-21T09:32:17.39663","exception":false,"start_time":"2023-02-21T09:32:17.379301","status":"completed"},"tags":[]},"source":["<a id=\"1.4\"></a>\n","## <b>1.4 <span style='color:#E1B12D'>Tuning hyperparameters in ensemble learning</span></b>"]},{"cell_type":"markdown","id":"88a56758","metadata":{"papermill":{"duration":0.0178,"end_time":"2023-02-21T09:32:17.432114","exception":false,"start_time":"2023-02-21T09:32:17.414314","status":"completed"},"tags":[]},"source":["In our case, predictions from one machine learning model become predictors for another (next level). Taking advantage of such complex composite structures can reduce prediction error. The final model becomes more robust, but high efficiency comes at the price of much more effort: much more time, computational resources and much more complex structure.\n","\n","Since the entire process is so complex we can think of it as a hyperparameters tuning strategy. This could be great material for different notebooks in the future!\n","\n","<div class=\"alert alert-block alert-info\" style=\"background-color:#3b3745;\n","                                                 color:white;\n","                                                 border-color:black;\n","                                                 border-radius:5px;\n","                                                 margin: auto;\n","                                                 text-align: left;\">\n","There are many different strategies, but we also have a very fundamental approach here. First all single models of the first level are trained and tuned (hyperparameters are tuned separately for each model), - then the ensemble model(s). During the next step the process is repeated in iterative movement from left to right.\n","</div>"]},{"cell_type":"markdown","id":"a7e69a44","metadata":{"papermill":{"duration":0.017289,"end_time":"2023-02-21T09:32:17.467151","exception":false,"start_time":"2023-02-21T09:32:17.449862","status":"completed"},"tags":[]},"source":["<a href=\"#toc\" role=\"button\" aria-pressed=\"true\" >🔝Back to Table of Contents🔝</a>"]},{"cell_type":"markdown","id":"ac7ad5b1","metadata":{"papermill":{"duration":0.022102,"end_time":"2023-02-21T09:32:17.506844","exception":false,"start_time":"2023-02-21T09:32:17.484742","status":"completed"},"tags":[]},"source":["<a id=\"2\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:170%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">2 <span style='color:#CDA63A'>|</span> Set-up</div>"]},{"cell_type":"markdown","id":"86c8873f","metadata":{"papermill":{"duration":0.025282,"end_time":"2023-02-21T09:32:17.551838","exception":false,"start_time":"2023-02-21T09:32:17.526556","status":"completed"},"tags":[]},"source":["<a id=\"2.1\"></a>\n","## <b>2.1 <span style='color:#E1B12D'>Import Libraries</span></b>"]},{"cell_type":"code","execution_count":3,"id":"722a5d09","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:17.590912Z","iopub.status.busy":"2023-02-21T09:32:17.590482Z","iopub.status.idle":"2023-02-21T09:32:18.873796Z","shell.execute_reply":"2023-02-21T09:32:18.872601Z"},"papermill":{"duration":1.305177,"end_time":"2023-02-21T09:32:18.876641","exception":false,"start_time":"2023-02-21T09:32:17.571464","status":"completed"},"tags":[]},"outputs":[],"source":["import numpy as np\n","import pandas as pd\n","import seaborn as sns\n","import plotly.express as px\n","import tkinter\n","from matplotlib import pyplot as plt\n","from sklearn.model_selection import cross_val_score\n","from collections import Counter"]},{"cell_type":"markdown","id":"12530b65","metadata":{"papermill":{"duration":0.017333,"end_time":"2023-02-21T09:32:18.911691","exception":false,"start_time":"2023-02-21T09:32:18.894358","status":"completed"},"tags":[]},"source":["<a id=\"2.2\"></a>\n","## <b>2.2 <span style='color:#E1B12D'>Import Data</span></b> "]},{"cell_type":"code","execution_count":4,"id":"bdf20666","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:18.947995Z","iopub.status.busy":"2023-02-21T09:32:18.94757Z","iopub.status.idle":"2023-02-21T09:32:19.021287Z","shell.execute_reply":"2023-02-21T09:32:19.02017Z"},"papermill":{"duration":0.094976,"end_time":"2023-02-21T09:32:19.023986","exception":false,"start_time":"2023-02-21T09:32:18.92901","status":"completed"},"tags":[]},"outputs":[],"source":["try:\n","    raw_df = pd.read_csv('/kaggle/input/telco-customer-churn/WA_Fn-UseC_-Telco-Customer-Churn.csv')\n","except:\n","    raw_df = pd.read_csv('telco.csv')"]},{"cell_type":"markdown","id":"8a9338a9","metadata":{"papermill":{"duration":0.017222,"end_time":"2023-02-21T09:32:19.058762","exception":false,"start_time":"2023-02-21T09:32:19.04154","status":"completed"},"tags":[]},"source":["<a id=\"2.3\"></a>\n","## <b>2.3 <span style='color:#E1B12D'>Data Set Characteristics</span></b>"]},{"cell_type":"markdown","id":"8220b469","metadata":{"papermill":{"duration":0.017063,"end_time":"2023-02-21T09:32:19.093366","exception":false,"start_time":"2023-02-21T09:32:19.076303","status":"completed"},"tags":[]},"source":["Each row represents a customer, each column contains customer’s attributes described on the column Metadata.\n","\n","The data set includes information about:\n","\n","* Customers who left within the last month – the column is called Churn.\n","* Services that each customer has signed up for – phone, multiple lines, internet, online security, online backup, device protection, tech support, and streaming TV and movies.\n","* Customer account information – how long they’ve been a customer, contract, payment method, paperless billing, monthly charges, and total charges.\n","* Demographic info about customers – gender, age range, and if they have partners and dependents.\n","\n","The are no missing values in data set."]},{"cell_type":"markdown","id":"68b10663","metadata":{"papermill":{"duration":0.017146,"end_time":"2023-02-21T09:32:19.127896","exception":false,"start_time":"2023-02-21T09:32:19.11075","status":"completed"},"tags":[]},"source":["<a id=\"2.4\"></a>\n","## <b>2.4 <span style='color:#E1B12D'>Dataset Attributes</span></b> "]},{"cell_type":"markdown","id":"d35dbf17","metadata":{"papermill":{"duration":0.01737,"end_time":"2023-02-21T09:32:19.163893","exception":false,"start_time":"2023-02-21T09:32:19.146523","status":"completed"},"tags":[]},"source":["\n","* <code>customerID</code> - Customer ID\n","\n","* <code>gender</code> - Whether the customer is a male or a female\n","\n","* <code>SeniorCitizen</code> - Whether the customer is a senior citizen (1, 0)\n","\n","* <code>Partner</code> - Whether the customer has a partner (Yes, No)\n","\n","* <code>Dependents</code> - Whether the customer has dependents (Yes, No)\n","\n","* <code>tenure</code> - Number of months the customer has stayed with the company\n","\n","* <code>PhoneService</code> - Whether the customer has a phone service (Yes, No)\n","\n","* <code>MultipleLines</code> - Whether the customer has multiple lines (Yes, No, No phone service)\n","\n","* <code>InternetService</code> - Customer’s internet service provider (DSL, Fiber optic, No)\n","\n","* <code>OnlineSecurity</code> - Whether the customer has online security (Yes, No, No internet service)\n","\n","* <code>OnlineBackup</code> - Whether the customer has online backup or not (Yes, No, No internet service)\n","\n","* <code>DeviceProtection</code> - Whether the customer has device protection (Yes, No, No internet service)\n","\n","* <code>TechSupport</code> - Whether the customer has tech support (Yes, No, No internet service)\n","\n","* <code>StreamingTV</code> - Whether the customer has streaming TV service (Yes, No, No internet service)\n","\n","* <code>StreamingMovies</code> - Whether the customer has streaming movies service (Yes, No, No internet service)\n","\n","* <code>Contract</code> - Indicates the type of the contract (Month-to-month, One year, Two year)\n","\n","* <code>PaperlessBilling</code> - Whether the customer has paperless billing (Yes, No)\n","\n","* <code>PaymentMethod</code> - Indicates the payment method (Electronic check, Mailed check, Bank transfer (automatic), Credit card (automatic))\n","\n","* <code>MonthlyCharges</code> - Indicates the current monthly subscription cost of the customer\n","\n","* <code>TotalCharges</code> - Indicates the total charges paid by the customer so far\n","\n","* <code>Churn</code> - Indicates whether the customer churned"]},{"cell_type":"code","execution_count":5,"id":"ce88a7de","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.200594Z","iopub.status.busy":"2023-02-21T09:32:19.200176Z","iopub.status.idle":"2023-02-21T09:32:19.205808Z","shell.execute_reply":"2023-02-21T09:32:19.204537Z"},"papermill":{"duration":0.02667,"end_time":"2023-02-21T09:32:19.208102","exception":false,"start_time":"2023-02-21T09:32:19.181432","status":"completed"},"tags":[]},"outputs":[],"source":["palette = ['#008080','#FF6347', '#E50000', '#D2691E'] # Creating color palette for plots"]},{"cell_type":"markdown","id":"2d275b6a","metadata":{"papermill":{"duration":0.017189,"end_time":"2023-02-21T09:32:19.242808","exception":false,"start_time":"2023-02-21T09:32:19.225619","status":"completed"},"tags":[]},"source":["<a href=\"#toc\" role=\"button\" aria-pressed=\"true\" >🔝Back to Table of Contents🔝</a>"]},{"cell_type":"markdown","id":"892e2039","metadata":{"papermill":{"duration":0.0172,"end_time":"2023-02-21T09:32:19.278172","exception":false,"start_time":"2023-02-21T09:32:19.260972","status":"completed"},"tags":[]},"source":["<a id=\"3\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:170%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">3 <span style='color:#CDA63A'>|</span> Data preprocessing</div>"]},{"cell_type":"markdown","id":"72507145","metadata":{"papermill":{"duration":0.017164,"end_time":"2023-02-21T09:32:19.312831","exception":false,"start_time":"2023-02-21T09:32:19.295667","status":"completed"},"tags":[]},"source":["We don't need a customerID column, so I'll delete it."]},{"cell_type":"code","execution_count":6,"id":"fa3b83c2","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.349949Z","iopub.status.busy":"2023-02-21T09:32:19.349224Z","iopub.status.idle":"2023-02-21T09:32:19.36451Z","shell.execute_reply":"2023-02-21T09:32:19.363396Z"},"papermill":{"duration":0.036654,"end_time":"2023-02-21T09:32:19.366935","exception":false,"start_time":"2023-02-21T09:32:19.330281","status":"completed"},"tags":[]},"outputs":[],"source":["df = raw_df.drop('customerID', axis=1)"]},{"cell_type":"markdown","id":"573a7283","metadata":{"papermill":{"duration":0.017386,"end_time":"2023-02-21T09:32:19.402065","exception":false,"start_time":"2023-02-21T09:32:19.384679","status":"completed"},"tags":[]},"source":["<a id=\"3.1\"></a>\n","## <b>3.1 <span style='color:#E1B12D'>Dealing with missing values in TotalCharges</span></b> "]},{"cell_type":"code","execution_count":7,"id":"bfa933e9","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.439212Z","iopub.status.busy":"2023-02-21T09:32:19.438746Z","iopub.status.idle":"2023-02-21T09:32:19.443796Z","shell.execute_reply":"2023-02-21T09:32:19.44257Z"},"papermill":{"duration":0.02688,"end_time":"2023-02-21T09:32:19.44646","exception":false,"start_time":"2023-02-21T09:32:19.41958","status":"completed"},"tags":[]},"outputs":[],"source":[" # df['TotalCharges'] = df['TotalCharges'].astype(float)"]},{"cell_type":"markdown","id":"fbd06b1e","metadata":{"papermill":{"duration":0.017362,"end_time":"2023-02-21T09:32:19.481668","exception":false,"start_time":"2023-02-21T09:32:19.464306","status":"completed"},"tags":[]},"source":["An error occured when trying to execule a code above: could not convert string to float: ''\n","\n","The string to be converted must not contain any characters or symbols. The error occurs due to the incorrect initialization of value to a string variable.\n","\n","We have probably empty strings in 'TotalCharges' colums, but as they were defined as string, they didn't appear as Null Values."]},{"cell_type":"code","execution_count":8,"id":"b76e447b","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.5232Z","iopub.status.busy":"2023-02-21T09:32:19.522363Z","iopub.status.idle":"2023-02-21T09:32:19.53564Z","shell.execute_reply":"2023-02-21T09:32:19.534372Z"},"papermill":{"duration":0.037683,"end_time":"2023-02-21T09:32:19.539295","exception":false,"start_time":"2023-02-21T09:32:19.501612","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["Number of entries with empty string:  11\n"]}],"source":["step1 = [len(i.split()) for i in df['TotalCharges']] # splitting individual elements of 'Total_Charges'\n","step2 = [i for i in range(len(step1)) if step1[i] != 1] # storing the index values of 'Total_Charges' where length is not equal to 1.\n","print('Number of entries with empty string: ', len(step2))"]},{"cell_type":"markdown","id":"a8e8f22c","metadata":{"papermill":{"duration":0.019289,"end_time":"2023-02-21T09:32:19.586653","exception":false,"start_time":"2023-02-21T09:32:19.567364","status":"completed"},"tags":[]},"source":["We coud try to impute the missing values by building a model for that reason or fill them using some often used values in that kind of situations e.g. mean, median or mode, but it's simpler (and probably better) to just drop the coresponding rows from a dataset as the numbers of empty string is very low (11)."]},{"cell_type":"code","execution_count":9,"id":"f3497432","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.629957Z","iopub.status.busy":"2023-02-21T09:32:19.628708Z","iopub.status.idle":"2023-02-21T09:32:19.640443Z","shell.execute_reply":"2023-02-21T09:32:19.639339Z"},"papermill":{"duration":0.034528,"end_time":"2023-02-21T09:32:19.643076","exception":false,"start_time":"2023-02-21T09:32:19.608548","status":"completed"},"tags":[]},"outputs":[],"source":["df = df.drop(step2, axis = 0).reset_index(drop=True) # Dropping rows with no values in 'Total_charges'"]},{"cell_type":"code","execution_count":10,"id":"7ccb3fa2","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.681268Z","iopub.status.busy":"2023-02-21T09:32:19.680514Z","iopub.status.idle":"2023-02-21T09:32:19.688613Z","shell.execute_reply":"2023-02-21T09:32:19.687765Z"},"papermill":{"duration":0.029873,"end_time":"2023-02-21T09:32:19.691018","exception":false,"start_time":"2023-02-21T09:32:19.661145","status":"completed"},"tags":[]},"outputs":[],"source":["df['TotalCharges'] = df['TotalCharges'].astype(float) # Finally we can convert string to float in 'Total_charges' column"]},{"cell_type":"markdown","id":"08ea511c","metadata":{"papermill":{"duration":0.017584,"end_time":"2023-02-21T09:32:19.726451","exception":false,"start_time":"2023-02-21T09:32:19.708867","status":"completed"},"tags":[]},"source":["<a id=\"3.2\"></a>\n","## <b>3.2 <span style='color:#E1B12D'>Dealing with duplicated values</span></b> "]},{"cell_type":"code","execution_count":11,"id":"28ee560b","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.764558Z","iopub.status.busy":"2023-02-21T09:32:19.763869Z","iopub.status.idle":"2023-02-21T09:32:19.788333Z","shell.execute_reply":"2023-02-21T09:32:19.786545Z"},"papermill":{"duration":0.046275,"end_time":"2023-02-21T09:32:19.791022","exception":false,"start_time":"2023-02-21T09:32:19.744747","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["Number of duplicated values in training dataset:  22\n"]}],"source":["print('Number of duplicated values in training dataset: ', df.duplicated().sum())"]},{"cell_type":"code","execution_count":12,"id":"241fe1a0","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.828977Z","iopub.status.busy":"2023-02-21T09:32:19.828552Z","iopub.status.idle":"2023-02-21T09:32:19.851624Z","shell.execute_reply":"2023-02-21T09:32:19.850104Z"},"papermill":{"duration":0.044859,"end_time":"2023-02-21T09:32:19.854123","exception":false,"start_time":"2023-02-21T09:32:19.809264","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["Duplicated values dropped succesfully\n","****************************************************************************************************\n"]}],"source":["df.drop_duplicates(inplace=True)\n","print(\"Duplicated values dropped succesfully\")\n","print(\"*\" * 100)"]},{"cell_type":"markdown","id":"8d1e0a09","metadata":{"papermill":{"duration":0.017725,"end_time":"2023-02-21T09:32:19.889985","exception":false,"start_time":"2023-02-21T09:32:19.87226","status":"completed"},"tags":[]},"source":["<a id=\"3.3\"></a>\n","## <b>3.3 <span style='color:#E1B12D'>Creating numerical and categorical lists</span></b> "]},{"cell_type":"code","execution_count":13,"id":"ec71efd8","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.92829Z","iopub.status.busy":"2023-02-21T09:32:19.927486Z","iopub.status.idle":"2023-02-21T09:32:19.947197Z","shell.execute_reply":"2023-02-21T09:32:19.946109Z"},"papermill":{"duration":0.042064,"end_time":"2023-02-21T09:32:19.950212","exception":false,"start_time":"2023-02-21T09:32:19.908148","status":"completed"},"tags":[]},"outputs":[],"source":["# distinction is based on the number of different values in the column\n","columns = list(df.columns)\n","\n","categoric_columns = []\n","numeric_columns = []\n","\n","for i in columns:\n","    if len(df[i].unique()) > 6:\n","        numeric_columns.append(i)\n","    else:\n","        categoric_columns.append(i)\n","\n","categoric_columns = categoric_columns[:-1] # Excluding target:'Churn'"]},{"cell_type":"code","execution_count":14,"id":"948e5fc7","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:19.988638Z","iopub.status.busy":"2023-02-21T09:32:19.987836Z","iopub.status.idle":"2023-02-21T09:32:19.993854Z","shell.execute_reply":"2023-02-21T09:32:19.992653Z"},"papermill":{"duration":0.02827,"end_time":"2023-02-21T09:32:19.996793","exception":false,"start_time":"2023-02-21T09:32:19.968523","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["Numerical fetures:  ['tenure', 'MonthlyCharges', 'TotalCharges']\n","Categorical fetures:  ['gender', 'SeniorCitizen', 'Partner', 'Dependents', 'PhoneService', 'MultipleLines', 'InternetService', 'OnlineSecurity', 'OnlineBackup', 'DeviceProtection', 'TechSupport', 'StreamingTV', 'StreamingMovies', 'Contract', 'PaperlessBilling', 'PaymentMethod']\n"]}],"source":["print('Numerical fetures: ',numeric_columns)\n","print('Categorical fetures: ',categoric_columns)"]},{"cell_type":"markdown","id":"d31a9176","metadata":{"papermill":{"duration":0.018291,"end_time":"2023-02-21T09:32:20.033588","exception":false,"start_time":"2023-02-21T09:32:20.015297","status":"completed"},"tags":[]},"source":["<a href=\"#toc\" role=\"button\" aria-pressed=\"true\" >🔝Back to Table of Contents🔝</a>"]},{"cell_type":"markdown","id":"c22c55d7","metadata":{"papermill":{"duration":0.018411,"end_time":"2023-02-21T09:32:20.070754","exception":false,"start_time":"2023-02-21T09:32:20.052343","status":"completed"},"tags":[]},"source":["<a id=\"4\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:150%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">4 <span style='color:#CDA63A'>|</span> Feature Engineering</div>"]},{"cell_type":"markdown","id":"6c3b652f","metadata":{"papermill":{"duration":0.018053,"end_time":"2023-02-21T09:32:20.107509","exception":false,"start_time":"2023-02-21T09:32:20.089456","status":"completed"},"tags":[]},"source":["<a id=\"4.1\"></a>\n","## <b>4.1 <span style='color:#E1B12D'>Train test split - stratified splitting</span></b>"]},{"cell_type":"markdown","id":"98b2e7e0","metadata":{"papermill":{"duration":0.017987,"end_time":"2023-02-21T09:32:20.143846","exception":false,"start_time":"2023-02-21T09:32:20.125859","status":"completed"},"tags":[]},"source":["Stratified splitting means that when you generate a training / validation dataset split, it will attempt to keep the same percentages of classes in each split.\n","\n","These dataset divisions are usually generated randomly according to a target variable. However, when doing so, the proportions of the target variable among the different splits can differ, especially in the case of small datasets."]},{"cell_type":"code","execution_count":15,"id":"cda82304","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.182921Z","iopub.status.busy":"2023-02-21T09:32:20.182033Z","iopub.status.idle":"2023-02-21T09:32:20.195612Z","shell.execute_reply":"2023-02-21T09:32:20.194567Z"},"papermill":{"duration":0.035879,"end_time":"2023-02-21T09:32:20.198163","exception":false,"start_time":"2023-02-21T09:32:20.162284","status":"completed"},"tags":[]},"outputs":[],"source":["from sklearn.preprocessing import LabelEncoder\n","le = LabelEncoder()\n","df[['Churn']] = df[['Churn']].apply(le.fit_transform) # Applying LabelEncoder to 'Churn' column\n","\n","X = df.drop('Churn', axis=1)\n","y = df['Churn']"]},{"cell_type":"code","execution_count":16,"id":"38b9b89c","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.237941Z","iopub.status.busy":"2023-02-21T09:32:20.237116Z","iopub.status.idle":"2023-02-21T09:32:20.251741Z","shell.execute_reply":"2023-02-21T09:32:20.250508Z"},"papermill":{"duration":0.037553,"end_time":"2023-02-21T09:32:20.254588","exception":false,"start_time":"2023-02-21T09:32:20.217035","status":"completed"},"tags":[]},"outputs":[],"source":["from sklearn.model_selection import train_test_split\n","X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, test_size = 0.25, random_state = 42)"]},{"cell_type":"markdown","id":"042f1301","metadata":{"papermill":{"duration":0.018206,"end_time":"2023-02-21T09:32:20.291471","exception":false,"start_time":"2023-02-21T09:32:20.273265","status":"completed"},"tags":[]},"source":["> Check this notebook for **<mark style=\"background-color:#E1B12D;color:white;border-radius:5px;opacity:1.0\">MUCH more information</mark>** **<mark style=\"background-color:#3b3745;color:white;border-radius:5px;opacity:1.0\">about dealing with imbalanced datasets</mark>**: [Best techniques and metrics for Imbalanced Dataset](https://www.kaggle.com/code/marcinrutecki/best-techniques-and-metrics-for-imbalanced-dataset)"]},{"cell_type":"markdown","id":"2320fbc9","metadata":{"papermill":{"duration":0.01855,"end_time":"2023-02-21T09:32:20.328966","exception":false,"start_time":"2023-02-21T09:32:20.310416","status":"completed"},"tags":[]},"source":["<a id=\"4.2\"></a>\n","## <b>4.2 <span style='color:#E1B12D'>Feature scaling</span></b>"]},{"cell_type":"code","execution_count":17,"id":"3466360b","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.368957Z","iopub.status.busy":"2023-02-21T09:32:20.368077Z","iopub.status.idle":"2023-02-21T09:32:20.386137Z","shell.execute_reply":"2023-02-21T09:32:20.384874Z"},"papermill":{"duration":0.041003,"end_time":"2023-02-21T09:32:20.38881","exception":false,"start_time":"2023-02-21T09:32:20.347807","status":"completed"},"tags":[]},"outputs":[{"data":{"text/plain":["array([[ 0.27610257,  1.01769998,  0.61931885],\n","       [-1.24037213, -1.53347618, -0.98780549],\n","       [-1.28135793, -1.33620306, -0.99803605],\n","       ...,\n","       [-1.19938633,  0.13498634, -0.90549797],\n","       [-0.05178385,  0.98259205,  0.30461881],\n","       [ 0.39905998,  1.17652156,  0.85307838]])"]},"execution_count":17,"metadata":{},"output_type":"execute_result"}],"source":["from sklearn.preprocessing import StandardScaler\n","Standard_Scaler = StandardScaler()\n","Standard_Scaler.fit_transform(X_train[numeric_columns])\n","Standard_Scaler.transform(X_test[numeric_columns])"]},{"cell_type":"markdown","id":"7862c2ee","metadata":{"papermill":{"duration":0.018641,"end_time":"2023-02-21T09:32:20.426511","exception":false,"start_time":"2023-02-21T09:32:20.40787","status":"completed"},"tags":[]},"source":["<a id=\"4.3\"></a>\n","## <b>4.3 <span style='color:#E1B12D'>One hot Encoding</span></b>"]},{"cell_type":"code","execution_count":18,"id":"0938da9c","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.46693Z","iopub.status.busy":"2023-02-21T09:32:20.46649Z","iopub.status.idle":"2023-02-21T09:32:20.471433Z","shell.execute_reply":"2023-02-21T09:32:20.470266Z"},"papermill":{"duration":0.029007,"end_time":"2023-02-21T09:32:20.475072","exception":false,"start_time":"2023-02-21T09:32:20.446065","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["['gender', 'SeniorCitizen', 'Partner', 'Dependents', 'PhoneService', 'MultipleLines', 'InternetService', 'OnlineSecurity', 'OnlineBackup', 'DeviceProtection', 'TechSupport', 'StreamingTV', 'StreamingMovies', 'Contract', 'PaperlessBilling', 'PaymentMethod']\n"]}],"source":["print(categoric_columns)"]},{"cell_type":"code","execution_count":19,"id":"2465b9ec","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.517545Z","iopub.status.busy":"2023-02-21T09:32:20.516972Z","iopub.status.idle":"2023-02-21T09:32:20.529147Z","shell.execute_reply":"2023-02-21T09:32:20.528273Z"},"papermill":{"duration":0.035624,"end_time":"2023-02-21T09:32:20.531388","exception":false,"start_time":"2023-02-21T09:32:20.495764","status":"completed"},"tags":[]},"outputs":[],"source":["from sklearn.preprocessing import OneHotEncoder\n","from sklearn.compose import make_column_transformer\n","# Encoding multiple columns. Unfortunately you cannot pass a list here so you need to copy-paste all printed categorical columns.\n","transformer = make_column_transformer(\n","    (OneHotEncoder(handle_unknown='ignore'), \n","     ['gender', 'SeniorCitizen', 'Partner', 'Dependents', \n","      'PhoneService', 'MultipleLines', 'InternetService', \n","      'OnlineSecurity', 'OnlineBackup', 'DeviceProtection', \n","      'TechSupport', 'StreamingTV', 'StreamingMovies', \n","      'Contract', 'PaperlessBilling', 'PaymentMethod']))"]},{"cell_type":"markdown","id":"321e2cfb","metadata":{"papermill":{"duration":0.01855,"end_time":"2023-02-21T09:32:20.569654","exception":false,"start_time":"2023-02-21T09:32:20.551104","status":"completed"},"tags":[]},"source":["<a id=\"4.3.1\"></a>\n","### <b>4.3.1 <span style='color:#E1B12D'>X_train encoding</span></b>"]},{"cell_type":"code","execution_count":20,"id":"5e3929ec","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.610333Z","iopub.status.busy":"2023-02-21T09:32:20.609265Z","iopub.status.idle":"2023-02-21T09:32:20.664766Z","shell.execute_reply":"2023-02-21T09:32:20.663635Z"},"papermill":{"duration":0.078962,"end_time":"2023-02-21T09:32:20.667573","exception":false,"start_time":"2023-02-21T09:32:20.588611","status":"completed"},"tags":[]},"outputs":[],"source":["# Transforming\n","transformed = transformer.fit_transform(X_train)\n","# Transformating back\n","transformed_df = pd.DataFrame(transformed, columns=transformer.get_feature_names_out())\n","# One-hot encoding removed an index. Let's put it back:\n","transformed_df.index = X_train.index\n","# Joining tables\n","X_train = pd.concat([X_train, transformed_df], axis=1)\n","# Dropping old categorical columns\n","X_train.drop(categoric_columns, axis=1, inplace=True)"]},{"cell_type":"markdown","id":"f8cb2fea","metadata":{"papermill":{"duration":0.019246,"end_time":"2023-02-21T09:32:20.70658","exception":false,"start_time":"2023-02-21T09:32:20.687334","status":"completed"},"tags":[]},"source":["<a id=\"4.3.2\"></a>\n","### <b>4.3.2 <span style='color:#E1B12D'>X_test encoding</span></b>"]},{"cell_type":"code","execution_count":21,"id":"362a3116","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.74752Z","iopub.status.busy":"2023-02-21T09:32:20.747073Z","iopub.status.idle":"2023-02-21T09:32:20.773427Z","shell.execute_reply":"2023-02-21T09:32:20.771376Z"},"papermill":{"duration":0.050775,"end_time":"2023-02-21T09:32:20.776798","exception":false,"start_time":"2023-02-21T09:32:20.726023","status":"completed"},"tags":[]},"outputs":[],"source":["# Transforming\n","transformed = transformer.transform(X_test)\n","# Transformating back\n","transformed_df = pd.DataFrame(transformed, columns=transformer.get_feature_names_out())\n","# One-hot encoding removed an index. Let's put it back:\n","transformed_df.index = X_test.index\n","# Joining tables\n","X_test = pd.concat([X_test, transformed_df], axis=1)\n","# Dropping old categorical columns\n","X_test.drop(categoric_columns, axis=1, inplace=True)"]},{"cell_type":"markdown","id":"1a4ed6c5","metadata":{"papermill":{"duration":0.019161,"end_time":"2023-02-21T09:32:20.817507","exception":false,"start_time":"2023-02-21T09:32:20.798346","status":"completed"},"tags":[]},"source":["<a href=\"#toc\" role=\"button\" aria-pressed=\"true\" >🔝Back to Table of Contents🔝</a>"]},{"cell_type":"markdown","id":"5d2e417b","metadata":{"papermill":{"duration":0.01891,"end_time":"2023-02-21T09:32:20.856016","exception":false,"start_time":"2023-02-21T09:32:20.837106","status":"completed"},"tags":[]},"source":["<a id=\"5\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:150%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">5 <span style='color:#CDA63A'>|</span> Model building</div>"]},{"cell_type":"code","execution_count":22,"id":"48fae4c4","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.896749Z","iopub.status.busy":"2023-02-21T09:32:20.896357Z","iopub.status.idle":"2023-02-21T09:32:20.901494Z","shell.execute_reply":"2023-02-21T09:32:20.900594Z"},"papermill":{"duration":0.028284,"end_time":"2023-02-21T09:32:20.903605","exception":false,"start_time":"2023-02-21T09:32:20.875321","status":"completed"},"tags":[]},"outputs":[],"source":["from sklearn.metrics import confusion_matrix, recall_score, precision_score, f1_score, accuracy_score, roc_auc_score\n","from sklearn.metrics import classification_report\n","from sklearn.metrics import plot_confusion_matrix, roc_auc_score, plot_roc_curve"]},{"cell_type":"markdown","id":"2612864d","metadata":{"papermill":{"duration":0.018558,"end_time":"2023-02-21T09:32:20.941179","exception":false,"start_time":"2023-02-21T09:32:20.922621","status":"completed"},"tags":[]},"source":["<a id=\"5.1\"></a>\n","## <b>5.1 <span style='color:#F1A424'>Level-0 with 28 Models</span></b>"]},{"cell_type":"code","execution_count":23,"id":"4988f1cd","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:20.980774Z","iopub.status.busy":"2023-02-21T09:32:20.980389Z","iopub.status.idle":"2023-02-21T09:32:21.530135Z","shell.execute_reply":"2023-02-21T09:32:21.528848Z"},"papermill":{"duration":0.572664,"end_time":"2023-02-21T09:32:21.53276","exception":false,"start_time":"2023-02-21T09:32:20.960096","status":"completed"},"tags":[]},"outputs":[],"source":["from sklearn.ensemble import AdaBoostClassifier\n","from sklearn.ensemble import BaggingClassifier\n","from sklearn.naive_bayes import BernoulliNB\n","from sklearn.calibration import CalibratedClassifierCV\n","from sklearn.naive_bayes import CategoricalNB\n","from sklearn.multioutput import ClassifierChain\n","from sklearn.naive_bayes import ComplementNB\n","from sklearn.tree import DecisionTreeClassifier\n","from sklearn.dummy import DummyClassifier\n","from sklearn.tree import ExtraTreeClassifier\n","from sklearn.ensemble import ExtraTreesClassifier\n","from sklearn.naive_bayes import GaussianNB\n","from sklearn.gaussian_process import GaussianProcessClassifier\n","from sklearn.ensemble import GradientBoostingClassifier\n","from sklearn.ensemble import HistGradientBoostingClassifier\n","from sklearn.neighbors import KNeighborsClassifier\n","from sklearn.semi_supervised import LabelPropagation\n","from sklearn.semi_supervised import LabelSpreading\n","from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n","from sklearn.svm import LinearSVC\n","from sklearn.linear_model import LogisticRegression\n","from sklearn.linear_model import LogisticRegressionCV\n","from sklearn.neural_network import MLPClassifier\n","from sklearn.multioutput import MultiOutputClassifier\n","from sklearn.naive_bayes import MultinomialNB\n","from sklearn.neighbors import NearestCentroid\n","from sklearn.svm import NuSVC\n","from sklearn.multiclass import OneVsOneClassifier\n","from sklearn.multiclass import OneVsRestClassifier\n","from sklearn.multiclass import OutputCodeClassifier\n","from sklearn.linear_model import PassiveAggressiveClassifier\n","from sklearn.linear_model import Perceptron\n","from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\n","from sklearn.neighbors import RadiusNeighborsClassifier\n","from sklearn.ensemble import RandomForestClassifier\n","from sklearn.linear_model import RidgeClassifier\n","from sklearn.linear_model import RidgeClassifierCV\n","from sklearn.linear_model import SGDClassifier\n","from sklearn.svm import SVC\n","from sklearn.ensemble import StackingClassifier\n","#from sklearn.ensemble import VotingClassifier\n","\n","from xgboost import XGBClassifier\n","from catboost import CatBoostClassifier"]},{"cell_type":"code","execution_count":24,"id":"b19701ec","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:21.574913Z","iopub.status.busy":"2023-02-21T09:32:21.574503Z","iopub.status.idle":"2023-02-21T09:32:21.592505Z","shell.execute_reply":"2023-02-21T09:32:21.5915Z"},"papermill":{"duration":0.040821,"end_time":"2023-02-21T09:32:21.59467","exception":false,"start_time":"2023-02-21T09:32:21.553849","status":"completed"},"tags":[]},"outputs":[],"source":["estimators = []\n","estimators.append(('AdaBoostClassifier', AdaBoostClassifier(random_state=13) ))\n","estimators.append(('Bagging Classifier', BaggingClassifier(random_state=13) ))\n","estimators.append(('Bernoulli NB', BernoulliNB() ))\n","estimators.append(('Decision Tree Classifier', DecisionTreeClassifier(random_state=13) ))\n","estimators.append(('Dummy Classifier', DummyClassifier(random_state=13) ))\n","estimators.append(('Extra Tree Classifier', ExtraTreeClassifier(random_state=13) ))\n","estimators.append(('Extra Trees Classifier', ExtraTreesClassifier(random_state=13) ))\n","estimators.append(('Gaussian NB', GaussianNB() ))\n","estimators.append(('Gaussian Process Classifier', GaussianProcessClassifier(random_state=13) ))\n","estimators.append(('Gradient Boosting Classifier', GradientBoostingClassifier(random_state=13) ))\n","estimators.append(('Hist Gradient Boosting Classifier', HistGradientBoostingClassifier(random_state=13) ))\n","estimators.append(('KNN', KNeighborsClassifier() ))\n","#estimators.append(('Label Propagation', LabelPropagation() ))\n","#estimators.append(('Label Spreading', LabelSpreading() ))\n","estimators.append(('LogisticRegression', LogisticRegression(max_iter=1000, random_state=13)))\n","estimators.append(('Logistic Regression CV', LogisticRegressionCV(max_iter=1000, random_state=13) ))\n","estimators.append(('MLPClassifier', MLPClassifier(max_iter=2000,random_state=13) ))\n","estimators.append(('Nearest Centroid', NearestCentroid() ))\n","estimators.append(('Passive Aggressive Classifier', PassiveAggressiveClassifier(random_state=13) ))\n","estimators.append(('Perceptron', Perceptron(random_state=13) ))\n","#estimators.append(('RadiusNeighborsClassifier', RadiusNeighborsClassifier(radius=3) ))\n","estimators.append(('RandomForest', RandomForestClassifier(max_depth= 10, min_samples_leaf= 1, min_samples_split= 3, n_estimators= 170, random_state=13) ))\n","estimators.append(('Ridge Classifier', RidgeClassifier(random_state=13) ))\n","estimators.append(('Ridge Classifier CV', RidgeClassifierCV() ))\n","estimators.append(('SGDClassifier', SGDClassifier(random_state=13) ))\n","estimators.append(('SVC', SVC(random_state=13)))\n","estimators.append(('XGB', XGBClassifier(random_state=13) ))\n","estimators.append(('CatBoost', CatBoostClassifier(logging_level='Silent', random_state=13) ))"]},{"cell_type":"markdown","id":"f661a339","metadata":{"papermill":{"duration":0.019316,"end_time":"2023-02-21T09:32:21.634112","exception":false,"start_time":"2023-02-21T09:32:21.614796","status":"completed"},"tags":[]},"source":["<a id=\"5.2\"></a>\n","## <b>5.2 <span style='color:#F1A424'>Level-1 Model</span></b>"]},{"cell_type":"code","execution_count":25,"id":"c7c74472","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:21.676297Z","iopub.status.busy":"2023-02-21T09:32:21.675838Z","iopub.status.idle":"2023-02-21T09:32:21.680283Z","shell.execute_reply":"2023-02-21T09:32:21.679447Z"},"papermill":{"duration":0.027938,"end_time":"2023-02-21T09:32:21.68233","exception":false,"start_time":"2023-02-21T09:32:21.654392","status":"completed"},"tags":[]},"outputs":[],"source":["XGB = XGBClassifier(random_state=13)"]},{"cell_type":"markdown","id":"2e1dc392","metadata":{"papermill":{"duration":0.019559,"end_time":"2023-02-21T09:32:21.721556","exception":false,"start_time":"2023-02-21T09:32:21.701997","status":"completed"},"tags":[]},"source":["<a id=\"5.3\"></a>\n","## <b>5.3 <span style='color:#F1A424'>Create a stacking classifier</span></b>"]},{"cell_type":"code","execution_count":26,"id":"5d112669","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:32:21.761534Z","iopub.status.busy":"2023-02-21T09:32:21.761153Z","iopub.status.idle":"2023-02-21T09:37:29.447751Z","shell.execute_reply":"2023-02-21T09:37:29.446386Z"},"papermill":{"duration":307.710408,"end_time":"2023-02-21T09:37:29.451052","exception":false,"start_time":"2023-02-21T09:32:21.740644","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["\n","Stacking classifier training Accuracy: 0.72\n","Stacking classifier test Accuracy: 0.77\n"]}],"source":["from sklearn.ensemble import StackingClassifier\n","SC = StackingClassifier(estimators=estimators,final_estimator=XGB,cv=6)\n","SC.fit(X_train, y_train)\n","y_pred = SC.predict(X_test)\n","\n","print(f\"\\nStacking classifier training Accuracy: {SC.score(X_train, y_train):0.2f}\")\n","print(f\"Stacking classifier test Accuracy: {SC.score(X_test, y_test):0.2f}\")"]},{"cell_type":"code","execution_count":27,"id":"20f6a04d","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:37:29.494107Z","iopub.status.busy":"2023-02-21T09:37:29.493679Z","iopub.status.idle":"2023-02-21T09:37:29.509023Z","shell.execute_reply":"2023-02-21T09:37:29.507607Z"},"papermill":{"duration":0.040191,"end_time":"2023-02-21T09:37:29.512013","exception":false,"start_time":"2023-02-21T09:37:29.471822","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["[[1096  193]\n"," [ 208  256]]\n"]}],"source":["SC_Recall = recall_score(y_test, y_pred)\n","SC_Precision = precision_score(y_test, y_pred)\n","SC_f1 = f1_score(y_test, y_pred)\n","SC_accuracy = accuracy_score(y_test, y_pred)\n","SC_roc_auc = roc_auc_score(y_test, y_pred)\n","\n","cm = confusion_matrix(y_test, y_pred)\n","print(cm)"]},{"cell_type":"code","execution_count":28,"id":"ec6f6321","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:37:29.555002Z","iopub.status.busy":"2023-02-21T09:37:29.554597Z","iopub.status.idle":"2023-02-21T09:54:14.354579Z","shell.execute_reply":"2023-02-21T09:54:14.353203Z"},"papermill":{"duration":1004.833593,"end_time":"2023-02-21T09:54:14.365936","exception":false,"start_time":"2023-02-21T09:37:29.532343","status":"completed"},"tags":[]},"outputs":[{"name":"stdout","output_type":"stream","text":["Cross Validation Recall scores are: [0.38351254 0.47311828 0.46402878 0.63669065 0.53046595]\n","Average Cross Validation Recall score:  0.49756323973079597\n","Cross Validation Recall standard deviation:  0.09376660013264607\n"]}],"source":["from statistics import stdev\n","score = cross_val_score(SC, X_train, y_train, cv=5, scoring='recall', error_score=\"raise\")\n","SC_cv_score = score.mean()\n","SC_cv_stdev = stdev(score)\n","print('Cross Validation Recall scores are: {}'.format(score))\n","print('Average Cross Validation Recall score: ', SC_cv_score)\n","print('Cross Validation Recall standard deviation: ', SC_cv_stdev)"]},{"cell_type":"code","execution_count":29,"id":"574ac3e5","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:54:14.407912Z","iopub.status.busy":"2023-02-21T09:54:14.407544Z","iopub.status.idle":"2023-02-21T09:54:14.426986Z","shell.execute_reply":"2023-02-21T09:54:14.425904Z"},"papermill":{"duration":0.043331,"end_time":"2023-02-21T09:54:14.429441","exception":false,"start_time":"2023-02-21T09:54:14.38611","status":"completed"},"tags":[]},"outputs":[{"data":{"text/html":["<div>\n","<style scoped>\n","    .dataframe tbody tr th:only-of-type {\n","        vertical-align: middle;\n","    }\n","\n","    .dataframe tbody tr th {\n","        vertical-align: top;\n","    }\n","\n","    .dataframe thead th {\n","        text-align: right;\n","    }\n","</style>\n","<table border=\"1\" class=\"dataframe\">\n","  <thead>\n","    <tr style=\"text-align: right;\">\n","      <th></th>\n","      <th>Model</th>\n","      <th>Recall</th>\n","      <th>Precision</th>\n","      <th>F1 Score</th>\n","      <th>Accuracy</th>\n","      <th>ROC-AUC Score</th>\n","      <th>Avg CV Recall</th>\n","      <th>Standard Deviation of CV Recall</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>Random Forest</td>\n","      <td>0.551724</td>\n","      <td>0.570156</td>\n","      <td>0.560789</td>\n","      <td>0.771249</td>\n","      <td>0.700998</td>\n","      <td>0.497563</td>\n","      <td>0.093767</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["           Model    Recall  Precision  F1 Score  Accuracy  ROC-AUC Score  \\\n","0  Random Forest  0.551724   0.570156  0.560789  0.771249       0.700998   \n","\n","   Avg CV Recall  Standard Deviation of CV Recall  \n","0       0.497563                         0.093767  "]},"execution_count":29,"metadata":{},"output_type":"execute_result"}],"source":["ndf = [(SC_Recall, SC_Precision, SC_f1, SC_accuracy, SC_roc_auc, SC_cv_score, SC_cv_stdev)]\n","\n","SC_score = pd.DataFrame(data = ndf, columns=['Recall','Precision','F1 Score', 'Accuracy', 'ROC-AUC Score', 'Avg CV Recall', 'Standard Deviation of CV Recall'])\n","SC_score.insert(0, 'Model', 'Random Forest')\n","SC_score"]},{"cell_type":"markdown","id":"a5d1aa6f","metadata":{"papermill":{"duration":0.01971,"end_time":"2023-02-21T09:54:14.469476","exception":false,"start_time":"2023-02-21T09:54:14.449766","status":"completed"},"tags":[]},"source":["<a id=\"5.4\"></a>\n","## <b>5.4 <span style='color:#F1A424'>AUC-ROC Curve</span></b>"]},{"cell_type":"code","execution_count":30,"id":"0745d8d1","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:54:14.510887Z","iopub.status.busy":"2023-02-21T09:54:14.510517Z","iopub.status.idle":"2023-02-21T09:54:20.147355Z","shell.execute_reply":"2023-02-21T09:54:20.145934Z"},"papermill":{"duration":5.661022,"end_time":"2023-02-21T09:54:20.15029","exception":false,"start_time":"2023-02-21T09:54:14.489268","status":"completed"},"tags":[]},"outputs":[{"data":{"image/png":"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\n","text/plain":["<Figure size 500x500 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["y_proba = SC.predict_proba(X_test)\n","\n","from sklearn.metrics import roc_curve\n","from sklearn.metrics import RocCurveDisplay\n","def plot_auc_roc_curve(y_test, y_pred):\n","    fpr, tpr, _ = roc_curve(y_test, y_pred)\n","    roc_display = RocCurveDisplay(fpr=fpr, tpr=tpr).plot()\n","    roc_display.figure_.set_size_inches(5,5)\n","    plt.plot([0, 1], [0, 1], color = 'g')\n","plot_auc_roc_curve(y_test, y_proba[:, 1])"]},{"cell_type":"markdown","id":"7c56987c","metadata":{"papermill":{"duration":0.020668,"end_time":"2023-02-21T09:54:20.193108","exception":false,"start_time":"2023-02-21T09:54:20.17244","status":"completed"},"tags":[]},"source":["<a id=\"5.5\"></a>\n","## <b>5.5 <span style='color:#F1A424'>Precision-Recall Curve (PR curve)</span></b>"]},{"cell_type":"markdown","id":"5722939a","metadata":{"papermill":{"duration":0.020971,"end_time":"2023-02-21T09:54:20.234859","exception":false,"start_time":"2023-02-21T09:54:20.213888","status":"completed"},"tags":[]},"source":["The precision-recall curve shows the tradeoff between precision and recalls for different thresholds."]},{"cell_type":"code","execution_count":31,"id":"4df6f970","metadata":{"execution":{"iopub.execute_input":"2023-02-21T09:54:20.279025Z","iopub.status.busy":"2023-02-21T09:54:20.278633Z","iopub.status.idle":"2023-02-21T09:54:26.450006Z","shell.execute_reply":"2023-02-21T09:54:26.448835Z"},"papermill":{"duration":6.19695,"end_time":"2023-02-21T09:54:26.452758","exception":false,"start_time":"2023-02-21T09:54:20.255808","status":"completed"},"tags":[]},"outputs":[{"data":{"image/png":"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\n","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["from sklearn.metrics import precision_recall_curve\n","from sklearn.metrics import PrecisionRecallDisplay\n","\n","display = PrecisionRecallDisplay.from_estimator(\n","    SC, X_test, y_test, name=\"Average precision\")\n","_ = display.ax_.set_title(\"Stacking Classifier\")"]},{"cell_type":"markdown","id":"699357d0","metadata":{"papermill":{"duration":0.022307,"end_time":"2023-02-21T09:54:26.496688","exception":false,"start_time":"2023-02-21T09:54:26.474381","status":"completed"},"tags":[]},"source":["<a id=\"5.6\"></a>\n","## <b>5.6 <span style='color:#F1A424'>This is just the beginning</span></b>"]},{"cell_type":"markdown","id":"1565aca1","metadata":{"papermill":{"duration":0.021903,"end_time":"2023-02-21T09:54:26.541029","exception":false,"start_time":"2023-02-21T09:54:26.519126","status":"completed"},"tags":[]},"source":["Creating the best model especcialt takeing advantage of stacking idea is a lengthy, iterative task!\n","\n","In our case, predictions from one machine learning model become predictors for another (next level). Taking advantage of such complex composite structures can reduce prediction error. The final model becomes more robust, but high efficiency comes at the price of much more effort: much more time, computational resources and much more complex structure.\n","\n","<div style=\"background-color:#CDA63A;\n","            color:white;\n","            text-align:left;\n","            width:70%;\n","            padding: 13px 13px; \n","            border-radius: 8px; \n","            margin: auto\">\n","Without tuning hyperparameters we are not making progress here!\n","</div>\n","\n","Since the entire process is so complex we can think of it as a hyperparameters tuning strategy. This could be great material for different notebooks in the future!\n","\n","<div class=\"alert alert-block alert-info\" style=\"background-color:#3b3745;\n","                                                 color:white;\n","                                                 border-color:black;\n","                                                 border-radius:5px;\n","                                                 margin: auto;\n","                                                 text-align: left;\">\n","There are many different strategies, but we also have a very fundamental approach here. First all single models of the first level are trained and tuned (hyperparameters are tuned separately for each model), - then the ensemble model(s). During the next step the process is repeated in iterative movement from left to right.\n","</div>"]},{"cell_type":"markdown","id":"13ffc792","metadata":{"papermill":{"duration":0.021888,"end_time":"2023-02-21T09:54:26.585236","exception":false,"start_time":"2023-02-21T09:54:26.563348","status":"completed"},"tags":[]},"source":["<a href=\"#toc\" role=\"button\" aria-pressed=\"true\" >🔝Back to Table of Contents🔝</a>"]},{"cell_type":"markdown","id":"5e4ea014","metadata":{"papermill":{"duration":0.022014,"end_time":"2023-02-21T09:54:26.629577","exception":false,"start_time":"2023-02-21T09:54:26.607563","status":"completed"},"tags":[]},"source":["<a id=\"6\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:150%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">5 <span style='color:#CDA63A'>|</span> My other notebooks</div>"]},{"cell_type":"markdown","id":"8941bf2d","metadata":{"papermill":{"duration":0.022297,"end_time":"2023-02-21T09:54:26.674465","exception":false,"start_time":"2023-02-21T09:54:26.652168","status":"completed"},"tags":[]},"source":["Here is a list of the notebooks I have completed so far:\n","\n","* [How to create a meaningful EDA](https://www.kaggle.com/code/marcinrutecki/how-to-create-a-meaningful-eda)\n","* [GridSearchCV + KFold CV: The Right Way](https://www.kaggle.com/code/marcinrutecki/gridsearchcv-kfold-cv-the-right-way)\n","* [Voting Classifier for Better Results](https://www.kaggle.com/code/marcinrutecki/voting-classifier-for-better-results)\n","* [SMOTE and Tomek Links for imbalanced data](https://www.kaggle.com/code/marcinrutecki/smote-and-tomek-links-for-imbalanced-data)\n","* [Best techniques and metrics for Imbalanced Dataset](https://www.kaggle.com/code/marcinrutecki/best-techniques-and-metrics-for-imbalanced-dataset)\n","* [Regression Models Evaluation Metrics](https://www.kaggle.com/code/marcinrutecki/regression-models-evaluation-metrics)\n","* [Clustering methods - comprehensive study](https://www.kaggle.com/code/marcinrutecki/clustering-methods-comprehensive-study)\n","* [Outlier detection methods!](https://www.kaggle.com/code/marcinrutecki/outlier-detection-methods)\n","* [Multicollinearity - detection and remedies](https://www.kaggle.com/code/marcinrutecki/multicollinearity-detection-and-remedies)"]},{"cell_type":"markdown","id":"6d2bcad0","metadata":{"papermill":{"duration":0.021972,"end_time":"2023-02-21T09:54:26.718891","exception":false,"start_time":"2023-02-21T09:54:26.696919","status":"completed"},"tags":[]},"source":["<a href=\"#toc\" role=\"button\" aria-pressed=\"true\" >🔝Back to Table of Contents🔝</a>"]},{"cell_type":"markdown","id":"e73e4aee","metadata":{"papermill":{"duration":0.021894,"end_time":"2023-02-21T09:54:26.7632","exception":false,"start_time":"2023-02-21T09:54:26.741306","status":"completed"},"tags":[]},"source":["<a id=\"7\"></a>\n","<div style=\"padding:20px;\n","            color:white;\n","            margin:10;\n","            font-size:150%;\n","            text-align:left;\n","            display:fill;\n","            border-radius:5px;\n","            background-color:#294B8E;\n","            overflow:hidden;\n","            font-weight:700\">7 <span style='color:#CDA63A'>|</span> References</div>"]},{"cell_type":"markdown","id":"476ec566","metadata":{"papermill":{"duration":0.021902,"end_time":"2023-02-21T09:54:26.807472","exception":false,"start_time":"2023-02-21T09:54:26.78557","status":"completed"},"tags":[]},"source":["http://rasbt.github.io/mlxtend/user_guide/classifier/StackingClassifier/\n","\n","https://machinelearningmastery.com/stacking-ensemble-machine-learning-with-python/"]}],"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":"3.7.12"},"papermill":{"default_parameters":{},"duration":1340.442264,"end_time":"2023-02-21T09:54:27.754679","environment_variables":{},"exception":null,"input_path":"__notebook__.ipynb","output_path":"__notebook__.ipynb","parameters":{},"start_time":"2023-02-21T09:32:07.312415","version":"2.3.4"}},"nbformat":4,"nbformat_minor":5}