MetaBox

Baseline Library

7 MetaBBO-RL optimizers, 1 MetaBBO-SL optimizer and 11 classic optimizers have been integrated into this platform. Choose one or more of them to be the baseline(s) to test the performance of your own optimizer.

Supported MetaBBO-RL optimizers:

Name	Year	Related paper
DE-DDQN	2019	Deep reinforcement learning based parameter control in differential evolution
QLPSO	2019	A reinforcement learning-based communication topology in particle swarm optimization
DEDQN	2021	Differential evolution with mixed mutation strategy based on deep reinforcement learning
LDE	2021	Learning Adaptive Differential Evolution Algorithm From Optimization Experiences by Policy Gradient
RL-PSO	2021	Employing reinforcement learning to enhance particle swarm optimization methods
RLEPSO	2022	RLEPSO:Reinforcement learning based Ensemble particle swarm optimizer✱
RL-HPSDE	2022	Differential evolution with hybrid parameters and mutation strategies based on reinforcement learning

Supported MetaBBO-SL optimizer:

Name	Year	Related paper
RNN-OI	2017	Learning to learn without gradient descent by gradient descent

Supported classic optimizers:

Name	Year	Related paper
PSO	1995	Particle swarm optimization
DE	1997	Differential Evolution – A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces
CMA-ES	2001	Completely Derandomized Self-Adaptation in Evolution Strategies
Bayesian Optimization	2014	Bayesian Optimization: Open source constrained global optimization tool for Python
GL-PSO	2015	Genetic Learning Particle Swarm Optimization
sDMS_PSO	2015	A Self-adaptive Dynamic Particle Swarm Optimizer
j21	2021	Self-adaptive Differential Evolution Algorithm with Population Size Reduction for Single Objective Bound-Constrained Optimization: Algorithm j21
MadDE	2021	Improving Differential Evolution through Bayesian Hyperparameter Optimization
SAHLPSO	2021	Self-Adaptive two roles hybrid learning strategies-based particle swarm optimization
NL_SHADE_LBC	2022	NL-SHADE-LBC algorithm with linear parameter adaptation bias change for CEC 2022 Numerical Optimization
Random Search	-	-

Note that Random Search performs uniformly random sampling to optimize the fitness.

For running commands, use the corresponding agent class name and corresponding optimizer class name to specify the algorithms:

MetaBBO-RL baselines:

Algorithm Name	Corresponding Agent Class	Corresponding Backbone Optimizer Class
DE-DDQN	DE_DDQN_Agent	DE_DDQN_Optimizer
QLPSO	QLPSO_Agent	QLPSO_Optimizer
DEDQN	DEDQN_Agent	DEDQN_Optimizer
LDE	LDE_Agent	LDE_Optimizer
RL-PSO	RL_PSO_Agent	RL_PSO_Optimizer
RLEPSO	RLEPSO_Agent	RLEPSO_Optimizer
RL-HPSDE	RL_HPSDE_Agent	RL_HPSDE_Optimizer

MetaBBO-SL baseline:

Algorithm Name	Corresponding Agent Class	Corresponding Backbone Optimizer Class
RNN-OI	L2L_Agent	L2L_Optimizer

classic baselines:

Algorithm Name	Corresponding Optimizer Class
PSO	DEAP_PSO
DE	DEAP_DE
CMA-ES	DEAP_CMAES
Bayesian Optimization	BayesianOptimizer
GL-PSO	GL_PSO
sDMS_PSO	sDMS_PSO
j21	JDE21
MadDE	MadDE
SAHLPSO	SAHLPSO
NL_SHADE_LBC	NL_SHADE_LBC
Random Search	Random_search

For all baselines, we display their control parameter settings in:

Classic: control_parameters_classic

Metabbo: control_parameters_metabbo

Baselines’ performance

To facilitate the observation of our baselines and related metrics, we tested our baselines on two levels of difficulty on three datasets. All data are provided in content.md.

This site is open source. Improve this page.