Emergent Real-World Robotic Skills via Unsupervised Off-Policy Reinforcement Learning.
Archit SharmaMichael AhnSergey LevineVikash KumarKarol HausmanShixiang GuPublished in: Robotics: Science and Systems (2020)
Keyphrases
- reinforcement learning
- real world
- real robot
- supervised learning
- data sets
- function approximation
- robotic systems
- case study
- wide range
- data driven
- unsupervised learning
- state space
- mobile robot
- neural network
- semi supervised
- machine learning
- markov decision processes
- optimal policy
- synthetic data
- multi agent
- real time
- labeled data
- genetic algorithm
- model free
- lifelong learning
- robot control
- unsupervised manner
- perception action