Skip to content

We propose a new, unsupervised, and adaptive Decision-Making framework called SIDM for Reinforcement Learning. This approach handles high complexity environments without manual intervention, and increases sample efficiency and policy effectiveness. site at https://ringbdstack.github.io/SIDM/

Notifications You must be signed in to change notification settings

RingBDStack/SIDM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

36 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SIDM

Repository for the paper Effective Reinforcement Learning Based on Structural Information Principles.

a novel and general Structural Information principles-based framework for effective Decision-Making, namely SIDM, approached from an information-theoretic perspective.

a general framework and can be flexibly integrated with various RL algorithms, including Deep RL (DRL), Hierarchical RL (HRL), and Multi-Agent RL (MARL). Extensive and empirical experiments are conducted on a visual gridworld, nine continuous control tasks, six bipedal robot tasks, and a StarCraft II micromanagement benchmark.

About

We propose a new, unsupervised, and adaptive Decision-Making framework called SIDM for Reinforcement Learning. This approach handles high complexity environments without manual intervention, and increases sample efficiency and policy effectiveness. site at https://ringbdstack.github.io/SIDM/

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages