Hyperparameter Optimization of Random Forest Algorithm to Enhance Performance Metric Evaluation of 5G Coverage Prediction

Abstract

Utilizing of 5G technology has become a major focus in the development of more advanced and efficient telecommunications networks. In this context, 5G coverage prediction becomes an important aspect in network planning to ensure optimal user experience. In this study, we explore the use of Random Forest algorithm to predict 5G coverage, with special emphasis on the hyperparameter optimization process to improve model performance. We conduct experiments with various hyperparameter combinations, including 'max_depth', 'max_features', 'min_samples_leaf', 'min_samples_split', and 'n_estimators', using hyperparameter optimization techniques. The results show that by paying attention to the optimal combination of hyperparameters, we managed to significantly improve the performance of the model. The optimized model produces a Minimum Root Mean Squared Error (RMSE) of 0.6, which is much better than the Random Forest model without hyperparameter optimization which has an RMSE of 1.14. The result of this study confirms the importance of the hyperparameter optimization process in improving the accuracy and consistency of the Random Forest model for 5G coverage prediction. The results have important implications in supporting the development of a successful 5G network infrastructure in the future.

Keywords

• hyperparameter optimization
• coverage prediction
• machine learning
• random forest
• rmse