Due to the rapid growth in the data storage and data processing demands, the energy consumption of data centers is becoming a key challenge. Live migration of Virtual Machine (VM) is commonly proposed to reduce energy consumption and to optimize resource usage in the literature although it comes with essential drawbacks such as migration cost and Service Level Agreement Violation (SLAV). A novel efficient resource management policy for virtualized data centers which optimizes the number of resources to meet requirements of dynamic workloads without migration is proposed in this paper. The proposed approach contains an experiment based prediction module. Energy-performance trade off relies on comparison of the predicted and the active number of servers periodically. The performance evaluations of the proposed approach are performed by using real-world workload traces from Planet Lab. The proposed algorithm is also compared to the other techniques proposed in the literature based on VM migration.