In most automation, we use ANN or RNN based algorithms. This results well but the prior information is what actions were previously taken by a human, this cannot be the only measure of learning a process we know humans learn everything with experience. And the most appropriate algorithm to learn like a human is Reinforcement Learning