Marcin Andrychowicz

imitation learning (3) reinforcement learning (3) policy learning (3) continuous control (2) off-policy learning (2) neural network optimization (1) domain randomization (1) value iteration (1) gradient descent (1) kullback-leibler divergence (1) reward function (1) hyperparameter tuning (1) robot manipulation (1) entropy regularization (1) learned optimization (1) expert demonstration (1) sparse reward (1) hindsight experience replay (1) adversarial imitation learning (1) vision-based control (1)