Automatic Parameter Optimization Using Genetic Algorithm in Deep Reinforcement Learning for Robotic Manipulation Tasks

Adarsh Sehgal; Nicholas Ward; Hung La; Sushil Louis

Automatic Parameter Optimization Using Genetic Algorithm in Deep Reinforcement Learning for Robotic Manipulation Tasks

Robotics 2022-11-03 v2

Authors: Adarsh Sehgal , Nicholas Ward , Hung La , Sushil Louis

Abstract

Learning agents can make use of Reinforcement Learning (RL) to decide their actions by using a reward function. However, the learning process is greatly influenced by the elect of values of the hyperparameters used in the learning algorithm. This work proposed a Deep Deterministic Policy Gradient (DDPG) and Hindsight Experience Replay (HER) based method, which makes use of the Genetic Algorithm (GA) to fine-tune the hyperparameters' values. This method (GA+DDPG+HER) experimented on six robotic manipulation tasks: FetchReach; FetchSlide; FetchPush; FetchPickAndPlace; DoorOpening; and AuboReach. Analysis of these results demonstrated a significant increase in performance and a decrease in learning time. Also, we compare and provide evidence that GA+DDPG+HER is better than the existing methods.

Keywords

genetic algorithm hierarchical reinforcement learning reinforcement learning

Cite

@article{arxiv.2204.03656,
  title  = {Automatic Parameter Optimization Using Genetic Algorithm in Deep Reinforcement Learning for Robotic Manipulation Tasks},
  author = {Adarsh Sehgal and Nicholas Ward and Hung La and Sushil Louis},
  journal= {arXiv preprint arXiv:2204.03656},
  year   = {2022}
}

Comments

I want to replace previous submission by this new submission with same title

Automatic Parameter Optimization Using Genetic Algorithm in Deep Reinforcement Learning for Robotic Manipulation Tasks

Abstract

Keywords

Cite

Comments

Related papers