Regularization
Entropy Regularization
Introduced by Mnih et al. in Asynchronous Methods for Deep Reinforcement LearningEntropy Regularization is a type of regularization used in reinforcement learning. For on-policy policy gradient based methods like A3C, the same mutual reinforcement behaviour leads to a highly-peaked $\pi\left(a\mid{s}\right)$ towards a few actions or action sequences, since it is easier for the actor and critic to overoptimise to a small portion of the environment. To reduce this problem, entropy regularization adds an entropy term to the loss to promote action diversity:
$$H(X) = -\sum\pi\left(x\right)\log\left(\pi\left(x\right)\right) $$
Image Credit: Wikipedia
Source: Asynchronous Methods for Deep Reinforcement Learning
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| Continuous Control | 25 | 11.06% |
| Atari Games | 22 | 9.73% |
| OpenAI Gym | 16 | 7.08% |
| Decision Making | 8 | 3.54% |
| Image Classification | 8 | 3.54% |
| Multi-agent Reinforcement Learning | 7 | 3.10% |
| Semantic Segmentation | 7 | 3.10% |
| Object Detection | 6 | 2.65% |
| Domain Adaptation | 6 | 2.65% |
Usage Over Time
Components
| Component | Type |
|
|---|---|---|
| 🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |
