New answers tagged rewards
0
votes
Why does shifting all the rewards have a different impact on the performance of the agent?
Our paper 'Exploit Reward Shifting in Value-Based DRL' answered this question.
In the case mentioned, using a negative shift will lead to explorative behaviors, therefore the DQN agent has better ...
Top 50 recent answers are included
Related Tags
rewards × 120reinforcement-learning × 112
deep-rl × 19
markov-decision-process × 14
reward-functions × 14
reward-shaping × 12
return × 11
q-learning × 9
dqn × 9
policy-gradients × 8
reward-design × 8
comparison × 7
deep-learning × 6
proximal-policy-optimization × 6
sutton-barto × 6
ai-design × 5
papers × 4
ddpg × 4
gym × 4
neural-networks × 3
machine-learning × 3
game-ai × 3
value-functions × 3
policies × 3
open-ai × 3