We investigate sparse representations for control in reinforcement learn...
Value-based approaches can be difficult to use in continuous action spac...
A variety of representation learning approaches have been investigated f...
Learning new representations of input observations in machine learning i...