The virtualization of operating systems and network functions results in a large amount of virtual machines (VM) that are hosted in data-center servers called physical machines (PM). This requires an efficient management and consolidation in order to use as few physical resources as possible to reduce the energy consumption. This paper addresses the VM placement problem with the aim of minimizing energy consumption as well as optimizing resource wastage in a data center. To achieve this goal, we propose an approach based on statistical methods to measure the similarity between resource vectors of PMs and VMs. The proposed strategy allows minimizing the wastage of resources by placing VMs on PMs while keeping the intra-workload balancing of the different resources. The energy consumption minimization is taken into account by reinforcing the selection of the target PM criterion as well as the growth rate of energy consumption. The energy efficiency and the optimization of resource wastage are demonstrated through simulation experiments. The obtained results show that our proposed IRB- VMP algorithm significantly outperforms the state-of-the-art FFD and BFD algorithms regarding the considered metrics.