Augmented Random Search Tutorial - How to Train Robots to Walk!
Автор: Skowster the Geek
Загружено: 2018-10-08
Просмотров: 6141
Learn one of the most advanced reinforcement learning algorithms to emerge in 2018, which has advanced the field of robotics by leaps and bounds, Augmented Random Search. Follow along with the coding tutorial and teach your own robot how to walk in less than an hour!
Correction: at [5:00] normalized = (inputs - observation_mean) / observation_sigma
Code for this tutorial:
https://github.com/colinskow/move37/t...
Original paper:
https://arxiv.org/abs/1803.07055
Dot product tutorial:
https://www.mathsisfun.com/algebra/ma...
Standard deviation and variance tutorial:
https://www.mathsisfun.com/data/stand...
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке: