From 250f87b93d186469b10b85c8d41210ce7ef504bf Mon Sep 17 00:00:00 2001 From: sven1977 Date: Tue, 15 Oct 2019 08:53:34 +0200 Subject: [PATCH] Add algo paper links to README. --- README.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index 74354de..e41a330 100644 --- a/README.md +++ b/README.md @@ -106,10 +106,10 @@ Running Atari experiments on a 24Gb machine with a memory that can keep 120k rec ### Currently implemented algorithms (stay tuned for more): -- **DQN2015** (2015) -- **DDDQN** (double/dueling/n-step/prioritized-replay Q-learning) -- **SAC** (Soft Actor-Critic in its Feb 2019 version) for continuous as well as discrete action spaces. -- Coming up next: **DADS** (an auto option discovery algo) +- **DQN2015** (deep Q-Networkin its 2015 Nature paper version) - [paper](https://www.nature.com/articles/nature14236) +- **DDDQN** (double/dueling/n-step/prioritized-replay Q-learning) - papers: [double](https://arxiv.org/abs/1509.06461) [dueling](https://arxiv.org/abs/1511.06581) [n-step](https://arxiv.org/abs/1901.07510) [PR](https://arxiv.org/abs/1511.05952) +- **SAC** (Soft Actor-Critic in its Feb 2019 version for continuous AND discrete(!) action spaces) - [paper](https://arxiv.org/abs/1812.05905) +- *Coming up next:* **DADS** (Dynamics-Aware Unsupervised Discovery of Skills) - [paper](https://arxiv.org/abs/1907.01657)