Robotics Posts - Page 3

Efficient reward learning is hard. With a focus on preference-based learning methods, we show how sample-efficiency can be achieved along with computational efficiency by using batch-active methods.