Machine learning ways are historically divided into 3 wide categories, which correspond to learning paradigms, according to the character with the "signal" or "opinions" accessible to the learning method: In reinforcement learning, the surroundings is often represented being a