In Mobile Ad Hoc wireless networks, communication devices establish a network on demand. This type of networks provides mobile communication capability to satisfy a need of a temporary nature and without the existence of any well-defined infrastructure. With the network topology changing dynamically, these networks tend to be vulnerable to several attacks. There is a need to devise security solutions to prevent attacks that affect the secure network operation. This paper proposes a novel way to detect wormhole attack using machine learning algorithms. The proposed algorithm utilized Ad hoc On-Demand Distance Vector (AODV) routing protocol to improve the detection method. Specially, we evaluate the performance of various machines learning classifier. The result indicates that Linear Discriminate Analysis gives high accuracy for simulated attack using NS3.