Optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.

Deep Reinforcement Learning (DRL) enables agents to make decisions based on a well-designed reward function that suites a particular environment without any prior knowledge related to a given environment. The adaptation of hyperparameters has a great impact on the overall learning process and the le...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Nesma M Ashraf, Reham R Mostafa, Rasha H Sakr, M Z Rashad
Formato: article
Lenguaje:EN
Publicado: Public Library of Science (PLoS) 2021
Materias:
R
Q
Acceso en línea:https://doaj.org/article/17aa2e0f666d4f07ba3b14861f6590be
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:17aa2e0f666d4f07ba3b14861f6590be
record_format dspace
spelling oai:doaj.org-article:17aa2e0f666d4f07ba3b14861f6590be2021-12-02T20:07:10ZOptimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.1932-620310.1371/journal.pone.0252754https://doaj.org/article/17aa2e0f666d4f07ba3b14861f6590be2021-01-01T00:00:00Zhttps://doi.org/10.1371/journal.pone.0252754https://doaj.org/toc/1932-6203Deep Reinforcement Learning (DRL) enables agents to make decisions based on a well-designed reward function that suites a particular environment without any prior knowledge related to a given environment. The adaptation of hyperparameters has a great impact on the overall learning process and the learning processing times. Hyperparameters should be accurately estimated while training DRL algorithms, which is one of the key challenges that we attempt to address. This paper employs a swarm-based optimization algorithm, namely the Whale Optimization Algorithm (WOA), for optimizing the hyperparameters of the Deep Deterministic Policy Gradient (DDPG) algorithm to achieve the optimum control strategy in an autonomous driving control problem. DDPG is capable of handling complex environments, which contain continuous spaces for actions. To evaluate the proposed algorithm, the Open Racing Car Simulator (TORCS), a realistic autonomous driving simulation environment, was chosen to its ease of design and implementation. Using TORCS, the DDPG agent with optimized hyperparameters was compared with a DDPG agent with reference hyperparameters. The experimental results showed that the DDPG's hyperparameters optimization leads to maximizing the total rewards, along with testing episodes and maintaining a stable driving policy.Nesma M AshrafReham R MostafaRasha H SakrM Z RashadPublic Library of Science (PLoS)articleMedicineRScienceQENPLoS ONE, Vol 16, Iss 6, p e0252754 (2021)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Nesma M Ashraf
Reham R Mostafa
Rasha H Sakr
M Z Rashad
Optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.
description Deep Reinforcement Learning (DRL) enables agents to make decisions based on a well-designed reward function that suites a particular environment without any prior knowledge related to a given environment. The adaptation of hyperparameters has a great impact on the overall learning process and the learning processing times. Hyperparameters should be accurately estimated while training DRL algorithms, which is one of the key challenges that we attempt to address. This paper employs a swarm-based optimization algorithm, namely the Whale Optimization Algorithm (WOA), for optimizing the hyperparameters of the Deep Deterministic Policy Gradient (DDPG) algorithm to achieve the optimum control strategy in an autonomous driving control problem. DDPG is capable of handling complex environments, which contain continuous spaces for actions. To evaluate the proposed algorithm, the Open Racing Car Simulator (TORCS), a realistic autonomous driving simulation environment, was chosen to its ease of design and implementation. Using TORCS, the DDPG agent with optimized hyperparameters was compared with a DDPG agent with reference hyperparameters. The experimental results showed that the DDPG's hyperparameters optimization leads to maximizing the total rewards, along with testing episodes and maintaining a stable driving policy.
format article
author Nesma M Ashraf
Reham R Mostafa
Rasha H Sakr
M Z Rashad
author_facet Nesma M Ashraf
Reham R Mostafa
Rasha H Sakr
M Z Rashad
author_sort Nesma M Ashraf
title Optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.
title_short Optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.
title_full Optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.
title_fullStr Optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.
title_full_unstemmed Optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.
title_sort optimizing hyperparameters of deep reinforcement learning for autonomous driving based on whale optimization algorithm.
publisher Public Library of Science (PLoS)
publishDate 2021
url https://doaj.org/article/17aa2e0f666d4f07ba3b14861f6590be
work_keys_str_mv AT nesmamashraf optimizinghyperparametersofdeepreinforcementlearningforautonomousdrivingbasedonwhaleoptimizationalgorithm
AT rehamrmostafa optimizinghyperparametersofdeepreinforcementlearningforautonomousdrivingbasedonwhaleoptimizationalgorithm
AT rashahsakr optimizinghyperparametersofdeepreinforcementlearningforautonomousdrivingbasedonwhaleoptimizationalgorithm
AT mzrashad optimizinghyperparametersofdeepreinforcementlearningforautonomousdrivingbasedonwhaleoptimizationalgorithm
_version_ 1718375286802743296