Agents - sankhaMukherjee/RLalgos GitHub Wiki
Agents comprise of the heart of DRL algorithms. There have been many types of algorithms that have been proposed, and these together form the group of algorithms that make what DRL is today. This repository comprises of a list of RL algorithms that can be used for generating different types of RL algorithms.
This will be a growing document. Currently the following algorithms have been implemented. Expect more to follow.
No. | Algorithm | Links | description |
---|---|---|---|
1. | DQN | Paper | API | Tutorial | This is the original DQN algorithm proposed by the DeepMind team |
2. | DoubleDQN | Paper | API | Tutorial | An improvement on the original wherein the Q value is not overestimated, as in the case of the original DQN algorithm |