List your company

Continuous Control on reacherSingle

1.01Average RewardDPG
0.979840.987670.99551.00333Sep 9, 2015

Evaluation results

MethodLinks
DPGSetup=original with replay buffer and batch normalization2015.091.011.083
DDPGObservation Space=pixels2015.0911.083
DDPGObservation Space=low-dimensional2015.090.9811.102