Learning-from-demonstrations is an emerging paradigm to obtain effectiverobot control policies for complex tasks via reinforcement learning without the need to explicitly design reward functions . We use Signal Temporal Logic to evaluate and rank the quality of demonstrations . We validate our approach through experiments on therete-world and OpenAI Gym environments . We show that our approachoutperforms the state-of-the-art Maximum Causal Entropy Inverse ReinforcementLearning . We also show that it performs the state of theart MaximumCausal Entropropiness Inverse Learning .

Author(s) : Aniruddh G. Puranic, Jyotirmoy V. Deshmukh, Stefanos Nikolaidis

Links : PDF - Abstract

Code :

https://github.com/mtrazzi/two-step-task


Coursera

Keywords : learning - demonstrations - state - logic - inverse -

Leave a Reply

Your email address will not be published. Required fields are marked *