Deep Reinforcement Learning using Genetic Algorithm for Parameter Optimization

Adarsh Sehgal, Hung Manh La, Sushil J. Louis, Hai Nguyen
2/19/2019
cs.NEcs.RO

Official code: aralab-unr/ReinforcementLearningWithGA (21 stars, MIT, last commit 2019-03-01).

Abstract

Reinforcement learning (RL) enables agents to take decision based on a reward function. However, in the process of learning, the choice of values for learning algorithm parameters can significantly impact the overall learning process. In this paper, we use a genetic algorithm (GA) to find the values of parameters used in Deep Deterministic Policy Gradient (DDPG) combined with Hindsight Experience Replay (HER), to help speed up the learning agent. We used this method on fetch-reach, slide, push, pick and place, and door opening in robotic manipulation tasks. Our experimental evaluation shows that our method leads to better performance, faster than the original algorithm.

View on arXivView PDF

Code Implementations(1)

aralab-unr/ReinforcementLearningWithGAOfficial100%
MIT
217HTML, PythonDec 5, 20187 years agoMIT

Similar Papers

Playing a 2D Game Indefinitely using NEAT and Reinforcement Learning

83% match2022

Evolutionary Reinforcement Learning: A Survey

82% match2023

Reinforcement Learning with Foundation Priors: Let the Embodied Agent Efficiently Learn on Its Own

81% match2023

Tactile Active Inference Reinforcement Learning for Efficient Robotic Manipulation Skill Acquisition

81% match2023

An Intelligent Social Learning-based Optimization Strategy for Black-box Robotic Control with Reinforcement Learning

80% match2023

Bigraph Matching Weighted with Learnt Incentive Function for Multi-Robot Task Allocation

80% match2024

Cite this paper

@article{sehgal2019deep,
  title  = {Deep Reinforcement Learning using Genetic Algorithm for Parameter Optimization},
  author = {Adarsh Sehgal and Hung Manh La and Sushil J. Louis and Hai Nguyen},
  year   = {2019},
  eprint = {1905.04100},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1905.04100}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1905.04100.svg)](https://codeofpaper.com/paper/1905.04100)

Discussion