Non-public networks (NPNs) have been considered as a promising technology to provide secure and customized services. There are several deployment options for NPNs according to the placement of network functions (NFs). To enhance the network performance, it is required to deploy more control plane NFs (CPFs) and user plane NFs (UPFs) in NPN. However, deploying more NFs can increase operating expenditure (OPEX) for the NPN operator. To assess this problem, this paper formulates a Markov decision process (MDP) and uses the policy iteration method to find the optimal policy. The simulation results show that the proposed scheme outperforms the comparison schemes in terms of the total reward.