Leveraging CAT Boost for Enhanced Prediction of App Ratings in the Google Play Store

This study focuses on predicting app ratings in the Google Play Store using a comprehensive google playstore dataset. It examines various app characteristics, such as the number of installs, app size, and the number of reviews, to determine influential factors in predicting app ratings. The methodology incorporates data preprocessing, feature engineering, and the application of five distinct regression models to evaluate their predictive performance. Among the models, CAT Boost emerged as the superior method, demonstrating the highest predictive accuracy. This research not only identifies key factors influencing app ratings but also offers a comparative analysis of model performances, facilitating the selection of the most effective predictive strategy. The insights gained are instrumental for developers seeking to enhance app quality and optimize user engagement.