How does Selecting a Benchmark Function Suite Influence the Estimation of an Algorithm's Quality?

2019 ◽  
Author(s):  
Iztok Fister ◽  
Suash Deb ◽  
Dusan Fister ◽  
Iztok Fister
Keyword(s):  
Benchmark Function

scholarly journals Real-Time Swarm Search Method for Real-World Quadcopter Drones

2018 ◽  
Vol 8 (7) ◽  
pp. 1169 ◽  
Author(s):  
Ki-Baek Lee ◽  
Young-Joo Kim ◽  
Young-Dae Hong
Keyword(s):  
Real Time ◽  
Optimization Algorithm ◽  
Real World ◽  
Particle Swarm Optimization Algorithm ◽  
Search Method ◽  
Air Pollutant ◽  
Function Optimization ◽  
Benchmark Function ◽  
World Environment ◽  
Repeated Test

This paper proposes a novel search method for a swarm of quadcopter drones. In the proposed method, inspired by the phenomena of swarms in nature, drones effectively look for the search target by investigating the evidence from the surroundings and communicating with each other. The position update mechanism is implemented using the particle swarm optimization algorithm as the swarm intelligence (a well-known swarm-based optimization algorithm), as well as a dynamic model for the drones to take the real-world environment into account. In addition, the mechanism is processed in real-time along with the movements of the drones. The effectiveness of the proposed method was verified through repeated test simulations, including a benchmark function optimization and air pollutant search problems. The results show that the proposed method is highly practical, accurate, and robust.

Parallel MCMC methods for global optimization

2019 ◽  
Vol 25 (3) ◽  
pp. 227-237
Author(s):  
Lihao Zhang ◽  
Zeyang Ye ◽  
Yuefan Deng
Keyword(s):  
Monte Carlo ◽  
Global Optimization ◽  
Statistical Theory ◽  
Simulated Annealing ◽  
Parallel Computers ◽  
Mcmc Methods ◽  
Benchmark Function ◽  
Mcmc Method ◽  
Validation And Verification ◽  
Parallel Scheme

Abstract We introduce a parallel scheme for simulated annealing, a widely used Markov chain Monte Carlo (MCMC) method for optimization. Our method is constructed and analyzed under the classical framework of MCMC. The benchmark function for optimization is used for validation and verification of the parallel scheme. The experimental results, along with the proof based on statistical theory, provide us with insights into the mechanics of the parallelization of simulated annealing for high parallel efficiency or scalability for large parallel computers.

A comparative study of teaching-learning-self-study algorithms on benchmark function optimization

2017 ◽  
Vol 34 (3) ◽  
pp. 628-641 ◽  
Author(s):  
Hyun-Jun Cho ◽  
Faisal Ahmed ◽  
Tae Young Kim ◽  
Beom Seok Kim ◽  
Yeong-Koo Yeo
Keyword(s):  
Comparative Study ◽  
Function Optimization ◽  
Benchmark Function ◽  
Self Study ◽  
Teaching Learning

scholarly journals Parametric Analysis of BFOA for Minimization Problems Using a Benchmark Function

Enfoque UTE ◽  
2019 ◽  
Vol 10 (3) ◽  
pp. 67-80
Author(s):  
Dannyll Michellc Zambrano Zambrano ◽  
Darío Vélez ◽  
Yohanna Daza ◽  
José Manuel Palomares
Keyword(s):  
Distributed Control ◽  
Parametric Analysis ◽  
Search Strategy ◽  
Benchmark Function ◽  
E Coli ◽  
Minimization Problems ◽  
The Social ◽  
Optimal Values ◽  
Simulation Results ◽  
Real World Problems

This paper presents the social foraging behavior of Escherichia coli (E. Coli) bacteria based on Bacteria Foraging Optimization algorithms (BFOA) to find optimization and distributed control values. The search strategy for E. coli is very complex to express and the dynamics of the simulated chemotaxis stage in BFOA is analyzed with the help of a simple mathematical model. The methodology starts from a detailed analysis of the parameters of bacterial swimming and tumbling (C) and the probability of elimination and dispersion (Ped), then an adaptive variant of BFOA is proposed, in which the size of the chemotherapeutic step is adjusted according to the current suitability of a virtual bacterium. To evaluate the performance of the algorithm in obtaining optimal values, the resolution was applied to one of the benchmark functions, in this case the Ackley minimization function, a comparative analysis of the BFOA is then performed. The simulation results have shown the validity of the optimal values (minimum or maximum) obtained on a specific function for real world problems, with a function belonging to the benchmark group of optimization functions.

Application of Dual-Adaptive Niched Genetic Algorithm in Optimal Design of Nuclear Power Components

2014 ◽  
Author(s):  
Cheng Wang ◽  
Chang-qi Yan ◽  
Jian-jun Wang ◽  
Lei Chen ◽  
Gui-jing Li
Keyword(s):  
Genetic Algorithm ◽  
Optimal Design ◽  
Nuclear Power ◽  
Optimization Design ◽  
Population Diversity ◽  
Geometric Constraints ◽  
Local Optimum ◽  
Penalty Function Method ◽  
Benchmark Function ◽  
Nonlinear Constraint

Genetic algorithm (GA) has been widely applied in optimal design of nuclear power components. Simple genetic algorithm (SGA) has the defects of poor convergence accuracy and easily falling into the local optimum when dealing with nonlinear constraint optimization problem. To overcome these defects, an improved genetic algorithm named dual-adaptive niched genetic algorithm (DANGA) is designed in this work. The new algorithm adopts niche technique to enhance global search ability, which utilizes a sharing function to maintain population diversity. Dual-adaptation technique is developed to improve the global and local search capability at the same time. Furthermore, a new reconstitution operator is applied to the DANGA to handle the constraint conditions, which can avoid the difficulty of selecting punishment parameter when using the penalty function method. The performance of new algorithm is evaluated by optimizing the benchmark function. The volume optimization of the Qinshan I steam generator and the weight optimization of Qinshan I condenser, taking thermal-hydraulic and geometric constraints into consideration, is carried out by adopting the DANGA. The result of benchmark function test shows that the new algorithm is more effective than some traditional genetic algorithms. The optimization design shows obvious validity and can provide guidance for real engineering design.

An Improved Particle Swarm Optimization Algorithm

2012 ◽  
Vol 195-196 ◽  
pp. 1060-1065
Author(s):  
Chang Yuan Jiang ◽  
Shu Guang Zhao ◽  
Li Zheng Guo ◽  
Chuan Ji
Keyword(s):  
Particle Swarm Optimization ◽  
Particle Swarm Optimization Algorithm ◽  
Particle Swarm ◽  
Pso Algorithm ◽  
Convergence Condition ◽  
Benchmark Function ◽  
Simulation Experiments ◽  
Swarm Optimization ◽  
Inertia Weight ◽  
Standard Particle Swarm Optimization

Based on the analyzing inertia weight of the standard particle swarm optimization (PSO) algorithm, an improved PSO algorithm is presented. Convergence condition of PSO is obtained through solving and analyzing the differential equation. By the experiments of four Benchmark function, the results show the performance of S-PSO improved more clearly than the standard PSO and random inertia weight PSO. Theoretical analysis and simulation experiments show that the S-PSO is efficient and feasible.

scholarly journals A Lightweight Learning Method for Stochastic Configuration Networks Using Non-Inverse Solution

Electronics ◽  
2022 ◽  
Vol 11 (2) ◽  
pp. 262
Author(s):  
Jing Nan ◽  
Zhonghua Jian ◽  
Chuanfeng Ning ◽  
Wei Dai
Keyword(s):  
Function Approximation ◽  
Inverse Method ◽  
Generalized Inverse ◽  
Low Complexity ◽  
High Complexity ◽  
Benchmark Function ◽  
Face Time ◽  
Generalized Inverse Method ◽  
Hidden Nodes ◽  
Real World Problems

Stochastic configuration networks (SCNs) face time-consuming issues when dealing with complex modeling tasks that usually require a mass of hidden nodes to build an enormous network. An important reason behind this issue is that SCNs always employ the Moore–Penrose generalized inverse method with high complexity to update the output weights in each increment. To tackle this problem, this paper proposes a lightweight SCNs, called L-SCNs. First, to avoid using the Moore–Penrose generalized inverse method, a positive definite equation is proposed to replace the over-determined equation, and the consistency of their solution is proved. Then, to reduce the complexity of calculating the output weight, a low complexity method based on Cholesky decomposition is proposed. The experimental results based on both the benchmark function approximation and real-world problems including regression and classification applications show that L-SCNs are sufficiently lightweight.

Sign in / Sign up

Export Citation Format

Share Document