Название: Deep Reinforcement Learning: Frontiers of Artificial Intelligence Автор: Mohit Sewak
Издательство: Springer
Год: 2019
Страниц: 215
Язык: английский
Формат: pdf (true)
Размер: 15.9 MB
Reinforcement Learning has evolved a long way with the enhancements from deep learning. Recent research efforts into combining deep learning with Reinforcement Learning have led to the development of some very powerful deep Reinforcement Learning systems, algorithms, and agents which have already achieved some extraordinary accomplishment. Not only have such systems surpassed the capabilities of most of the classical and non-deep-learning-based Reinforcement Learning agents, but have also started outperforming the best of human intelligence at tasks which were believed to require extreme human intelligence, creativity, and planning skills. Some of the DQN-based agents consistently beating the best of human players at the complex game of AlphaGo are very good examples of this.
