Code Snippet: Actor-critic on the Pendulum problem with continuous action
Description: Actor-critic on the Pendulum problem with continuous action, tile coding,
and a policy structure using a parameterized normal distribution.
Source code:
- Doxygen: ActorCriticPendulum.java
- Github: ActorCriticPendulum.java
Reference:
Model-Free Reinforcement Learning with Continuous Action in Practice. T. Degris, P. M. Pilarski, R. S. Sutton (2012). In Proceedings of the 2012 American Control Conference.Running this demo:
- From the command line:
- Download rlpark.jar
- Run the following command line:
java -cp rlpark.jar rlpark.example.demos.learning.ActorCriticPendulum
- In Zephyr standalone application:
- Download Zephyr standalone application
- Install RLPark plug-ins in Zephyr
- Go to:
- In Eclipse, as a Java Application:
- Create a new or use an existing project
- Include rlpark.jar in the project classpath
- Run a
Java Application
target usingrlpark.example.demos.learning.ActorCriticPendulum
as a main class
- In Eclipse, as an Eclipse Application:
- Install Zephyr plug-ins and
RLPark plug-ins in Eclipse
or
download RLPark source code and import RLPark projects (including the demo project rlpark.example.demos) into the workspace - Set up an
Eclipse Application
target following the tutorial Using Zephyr plug-ins - In the
Eclipse Application
target configuration:- In the menu, go to:
- Select the
Eclipse Application
target - In the rlpark.example.demos and rlpark.plugin.rltoysview to enable RLPark views tab, select the plug-in
- Start Zephyr by running the
Eclipse Application
target - In the Zephyr menu, go to:
or in the tab, addrlpark.example.demos.learning.ActorCriticPendulum
to the text field
- Install Zephyr plug-ins and
RLPark plug-ins in Eclipse
Dependencies
zephyr.plugin.core.api, rlpark.plugin.rltoysDocumentation