Optimizing resource allocation is crucial for quality of service if we are to maintain latency low, packet loss modest, and jitter reduced in 5G networks. The Random Forest Regressor may be used to estimate resource allocation requests by means of patterns acquired from traffic and network activity. Using important criteria like packet loss and jitter, the model's predictions are maximized as performance measures. As 5G wireless networks grow, ensuring outstanding QoS remains a critical challenge given the distinct and demanding demands of many applications. This paper offers a strategy using resource allocation based on Random Forest Regressor model with hyperparameter modification to enhance 5G network QoS. As said, applying the predictive powers of machine learning helps to maximize resource allocation, therefore improving the general network performance. Based on the experimental results, the proposed approach enhances management of network resources as well as the quality-of-service measurements including latency, dependability, throughput, packet loss, and jitter.