[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-127790-en":3,"doc-seo-127790-105":30,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},127790,1099523885074,"Ivy","https://ap-avatar.wpscdn.com/davatar_9964176cb1d06d4a9deccf72a44ae3dc",8,"Research & Report","Optimising Customer Churn Prediction in Telecommunications - A Comparative Analysis of Machine Learning Models Using Hierarchical Time Series Forecasting","Customer churn is a critical challenge in the telecommunications industry, where retaining existing customers is more cost-effective than acquiring new ones. This thesis evaluates SARIMAX, Prophet, ETS, LSTM, and XGBoost for predicting customer churn, extending beyond individual churner prediction. It applies hierarchical time series forecasting to estimate daily churn counts across multiple aggregation levels, enabling actionable resource-allocation insights. Results show Prophet and XGBoost achieve the strongest accuracy, while ETS and LSTM weaken at higher aggregation levels.","Master’s Thesis 2024 30 ECTS  \nSchool of Economics and Business (HH)  \nOptimising Customer Churn  \nPrediction in Telecommunications: A Comparative Analysis of Machine Learning Models Using Hierarchical Time Series Forecasting  \nAlin Dak Al-Bab  \nData Science  \nAbstract  \nCustomer churn is a critical challenge in the telecommunications industry, where retaining existing customers is more cost-effective than acquiring new ones. This thesis evaluates the effectiveness of machine learning models—SARIMAX, Prophet, ETS, LSTM, and XGBoost—in predicting customer churn. Unlike existing research that focuses on predicting individual churners, this study employs hierarchical time series forecasting to predict daily churn counts across various data aggregation levels, providing strategic insights for better resource allocation and customer retention.  \nUsing a dataset from 2017 to 2022, structured into hierarchical levels based on sales channels, age groups, and product types, key findings reveal that Prophet and XGBoost outperform other models in MSE, RMSE, MAE, and R2, with Prophet being the most consistent across levels. SARIMAX shows competitive performance at lower aggregation levels but is outperformed by Prophet and XGBoost at higher levels. ETS and LSTM models demonstrate poorer performance, especially in higher aggregation levels.  \nThis study identifies significant gaps in existing literature, particularly the limited use of hierarchical forecasting in churn prediction. By shifting the focus to churn count forecasting, this research provides a more comprehensive understanding of churn dynamics. The findings suggest that hierarchical forecasting models offer more accurate, actionable insights, enhancing strategic decision-making in customer retention.  \nThe thesis concludes with recommendations for future research, emphasising the potential for incorporating additional data sources and improving model interpretability. It demonstrates the effectiveness of hierarchical time series forecasting in predicting customer churn and provides valuable insights for the telecommunications industry.  \nFurthermore, this thesis contributes to the body of knowledge by showcasing the efficacy of hierarchical time series forecasting in churn prediction, offering valuable insights for the telecommunications industry, and setting the stage for future advancements in churn prediction methodologies.  \nTable of Contents  \nABSTRACT.................................................................................................................................................. I  \nLIST OF FIGURES .................................................................................................................................. IV  \nLIST OF TABLES ...................................................................................................................................... V  \n1. INTRODUCTION...............................................................................................................................1  \n1.1. INTRODUCTION AND BACKGROUND............................................................................................. 1  \n1.2. PROBLEM STATEMENT .................................................................................................................3  \n1.3. RESEARCH OBJECTIVES................................................................................................................4  \n1.4. RESEARCH QUESTIONS .................................................................................................................4  \n1.5. SIGNIFICANCE OF THE STUDY ......................................................................................................5  \n1.6. STRUCTURE OF THE THESIS ..........................................................................................................6  \n1.7. RESEARCH LIMITATIONS ................................................................................................","cbCaiaIAV2lbW1ny","https://ap.wps.com/l/cbCaiaIAV2lbW1ny","pdf",3801553,1,108,"English","en",105,"# Abstract\n# Introduction\n## Introduction and Background\n## Problem Statement\n## Research Objectives\n## Research Questions\n## Significance of the Study\n## Structure of the Thesis\n## Research Limitations\n## Data Privacy and Ethical Considerations\n# Literature Review\n## Customer Churn\n## Machine Learning Techniques for Churn Prediction\n## Hierarchical and Aggregated Time Series Forecasting\n## Review of Predictive Models in Churn Analysis\n## Gaps in Existing Research\n# Research Methodology\n## Introduction\n## Research Design\n## Data Collection and Selection\n## Data Collection\n## Data Selection","[{\"question\":\"Which machine learning models are evaluated for customer churn prediction?\",\"answer\":\"The thesis evaluates SARIMAX, Prophet, ETS, LSTM, and XGBoost to predict customer churn in telecommunications.\"},{\"question\":\"How does the study differ from existing churn-prediction research?\",\"answer\":\"Instead of focusing on predicting individual churners, it forecasts daily churn counts using hierarchical time series forecasting across multiple aggregation levels.\"},{\"question\":\"Which models perform best, and where do models struggle?\",\"answer\":\"Prophet and XGBoost outperform the others across accuracy metrics, with Prophet showing the most consistent results. ETS and LSTM show poorer performance, especially at higher aggregation levels.\"}]","Optimising Customer Churn Prediction in Telecommunications - A Comparative Analysis of Machine Learning Models Using Hierarchical Time Series Forecasting | PDF",1785941693,272,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"optimising-customer-churn-prediction-in-telecommunications-a-comparative-analysis-of-machine-learning-models-using-hierarchical-time-series-forecasting","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/optimising-customer-churn-prediction-in-telecommunications-a-comparative-analysis-of-machine-learning-models-using-hierarchical-time-series-forecasting/127790/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-22","2026-08-05",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Which machine learning models are evaluated for customer churn prediction?","Question",{"text":76,"@type":77},"The thesis evaluates SARIMAX, Prophet, ETS, LSTM, and XGBoost to predict customer churn in telecommunications.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the study differ from existing churn-prediction research?",{"text":81,"@type":77},"Instead of focusing on predicting individual churners, it forecasts daily churn counts using hierarchical time series forecasting across multiple aggregation levels.",{"name":83,"@type":74,"acceptedAnswer":84},"Which models perform best, and where do models struggle?",{"text":85,"@type":77},"Prophet and XGBoost outperform the others across accuracy metrics, with Prophet showing the most consistent results. ETS and LSTM show poorer performance, especially at higher aggregation levels.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]