In the past few years, an evolution in the wireless communication has been emerged, along with the evolution a new type with large potential application of wireless network appears, which is the Mobile Ad-Hoc Network (MANET). Black hole attack consider one of the most affected kind on MANET. Therefore, the use of intrusion detection system (IDS) has a major importance in the MANET protection. In this paper, an optimization of a fuzzy based intrusion detection system is proposed which automate the process of producing a fuzzy system by using an Adaptive Neuro-Fuzzy Inference System (ANFIS) for the initialization of the FIS and then optimize this initialized system by using Genetic Algorithm (GA). In addition, a normal estimated fuzzy based IDS is introduces to see the effect of the optimization on the system. From this study, it is proven that the optimized proposed IDS perform better that the normal estimated systems.