<p>5G has boosted the possibility of ultra-high-speed, low-latency, and reliable wireless communication systems. With 5G networks, if efficient resource management is not properly looked at, then the full potential of such networking cannot be harnessed. This work proposes a new hybrid machine learning based framework for dynamic resource management in a 5G network based on a combination of deep and reinforcement learning. This model allocates resources with regard to the current state of the network and users' requirements to increase the level of network performance during usage. It includes a Deep Neural Network (DNN) as a feature learner and Reinforcement Learning (RL) as a decision maker. The DNN learns different levels of data that are associated with the network, including network performance, user behavior, and environmental data, in an attempt to identify features that define 5G network dynamics. The RL agent in contact with the 5G network environment acquires the know-how for resource allocation in order to maximize a reward signal derived from the network performance measures such as latency, throughput, and energy consumption. Due to variations in resource allocation, the RL agent enhances the network's efficiency over time. The strength of the proposed model is found in the DL part for capturing highly nonlinear patterns in the network data and the RL agent's flexibility to operate in real-time. The analytical and experimental studies prove the model's potency in attaining intelligent and efficient management of the resources, thus making it a potential solution for 5G network management.</p>