Determine the optimal Hidden Layers and Neurons in the Generative Adversarial Networks topology for the Intrusion Detection Systems