Skip to content

This project implements Continuous Action Learning Automata (CALA) to learn an arbitrary set of actions for n-states.

Notifications You must be signed in to change notification settings

tjards/continuous_action_learning_automata

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Implementaton of Continuous Action Learning Automata (CALA)

This project implements Continuous Action Learning Automata (CALA) to learn an arbitrary set of actions for n-states. CALA is an on-policy Reinforcement Learning technique popular for solving stochastic optimization problems. For detailed formulations, see Reference 1.

Results

Below is a plot demonstrating performance for an n=3 state space with a learning rate of 0.1. Actions are constrained between -1 and 1. Note that the means converge to the correct values, variance decreases over time, and the rewards increase.

Results of n=3 State Space

References

  1. Beigy, H., Meybodi, M.R. (2003). A New Continuous Action-Set Learning Automaton for Function Optimization. In: Yazıcı, A., Şener, C. (eds) Computer and Information Sciences - ISCIS 2003. ISCIS 2003. Lecture Notes in Computer Science, vol 2869. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-39737-3_119

About

This project implements Continuous Action Learning Automata (CALA) to learn an arbitrary set of actions for n-states.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages