Purposive behavior acquisition for a real robot by vision-based reinforcement learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Reinforcement learning in robotics: A survey (2013, methodology): “A sequence of increasingly difficult missions has been employed to learn a goal shooting task in (Asada et al., 1996) using Q-learning. Randløv and Alstrøm (1998) discuss shaping the reward function to include both a balancing and a goal oriented term for a simulated bicycle riding task.”
Reverse Curriculum Generation for Reinforcement Learning (2017, methodology): “ currently considers too hard [13,14]. This type of adaptive curriculum has mainly been applied to supervised tasks, and most practical curriculum approaches in RL rely on pre-speciﬁed task sequences [15,16]. Some very general frameworks have been proposed to generate increasingly hard problems [17,”
Growing Action Spaces (2019, methodology): “With a resettable simulator, one can use a sequence of progressively more challenging initial states (Asada et al., 1996; Florensa et al., 2017).”
The Playground Experiment: Task-Independent Development of a Curious Robot (2005, unknown): ““Learning form easy mission” (( Asada et al. 1996 )).”