In this talk, we want to walk through two machine learning scopes as supervised and unsupervised learning. Most part of this talk will be focused on the second one, and when the minds are boiled, we will try to move around adversarial training with Generative Adversarial Networks. After that, we will be focused on different research ideas for using this network structure in different applications and I hope we will cover some ideas in implementing, stabilizing and different important factors to train adversarial networks.