Resource management plays the vital role in the cloud computing as the requirement for the massive data processing system such as heath sectors, business solutions and the internet of things keeps on increasing in at an exponential range. Allocation of proper and perfect resources remains as the mains reasons for the successful computation of the applications. However the conventional resources management methodologies, that totally depends on the simple heuristic based methods fails to accomplish a performance that is predictable. The appropriate resource allocation is directly related to the workload demand prediction as the would help to bring down the cost, time and power and the memory usage. The proposed method in the paper leverages the machine learning approaches to manage the resource allocation in the cloud computing for the massive data processing system, the simulation of the proposed model using the network simulator -2 enables to achieve a better performance and resources utilization at a decreased cost, time, power and memory usage.