Reinforcement Learning - Octopus Arm

Home

Octopus Arm

Octopus arm problem is for learning to control the simulated octopus arm through viscous environment to achieve a certain goal. By controlling each muscle in individual compartment, the agent moves the tip of the arm close to the goal. This is very interesting problem because of its high-dimensional actions, non-linear dynamics, and large search space.

Octopus Arm Simulator

Source code in R Change logs reference manual

Experimental data

Data collection tool in R

Sample data: 5000 steps of random actions (910K)

Sample data: 1000 steps of random actions (184K)

Contributors

Reinaldo Uribe (muriel): the simulator and data collecting tool

Last updated April 8, 2010