Search academic texts

Information × Registration Number 2124U004708, Article popup.category Препринт Title popup.author Sadova Oksana popup.publication 01-01-2024 popup.source_user Український католицький університет popup.source https://hdl.handle.net/20.500.14570/4837 popup.publisher Description User behavior and revenue forecasting are important aspects of business operations such as budgeting, financial management, and customer spend analysis. Focusing on the competitive and fast-growing online gaming industry, this thesis examines the unique challenges of growing a business in this sector. The performance of common machine learning algorithms for time series fore- casting, including SARIMAX, Random Forest, XGBoost, LSTM, and Facebook Prophet, is evaluated by applying them to predict user spending on an online gaming plat- form operated by a particular client. This thesis evaluates and compares the pre- diction accuracy of these models using measures such as mean absolute percentage error (MAPE), normalized root mean square error (NRMSE), and visual analysis. Next, the best methodologies for forecasting the demand and income of the com- pany under study are highlighted, and the results of the most effective algorithm are interpreted. After that, GGR forecasts for the next month are provided, and how these forecasts can be strategically used in business planning. popup.nrat_date 2025-05-09 Close