Simulation

Using Generative Adversarial Networks to Develop a Realistic Human Behavior Simulator

Simulation environments have proven to be very useful as testbeds for reinforcement learning (RL) algorithms. For settings where an actual human user is involved, these simulation environments allow one to test out the suitability of new RL …