Deep Reinforcement Learning is an approach in solving reinforcement learning problems that utilizes neural networks to represent any component of the agent like the policy, model, or value function that can learn and represent knowledge better and better by utilizing more resources, computation, data and experience from the environment.