Association rule mining based parameter adaptive strategy for differential evolution algorithms

Wang, Chuan; Liu, Yancheng; Zhang, Qinjin; Guo, Haohao; Liang, Xiaoling; Yang, Chen; Xu, Minyi; Wei, Yi

doi:10.1016/j.eswa.2019.01.035

Cited by 24 publications

(6 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Without prior knowledge, the number of hidden layer neurons is not easy to determine ( Wang et al, 2019 ). In order to reduce the error caused by the randomness of the neural network, this paper adopted the trial-and-error method ( Kermanshahi, 1998 ) to determine the number of hidden layer neurons.…”

Section: Experiments On Common Classification and Regression Problemsmentioning

confidence: 99%

An RBF neural network based on improved black widow optimization algorithm for classification and regression problems

Liu

Zhou²,

Zhou³

et al. 2023

Front. Neuroinform.

View full text Add to dashboard Cite

IntroductionRegression and classification are two of the most fundamental and significant areas of machine learning.MethodsIn this paper, a radial basis function neural network (RBFNN) based on an improved black widow optimization algorithm (IBWO) has been developed, which is called the IBWO-RBF model. In order to enhance the generalization ability of the IBWO-RBF neural network, the algorithm is designed with nonlinear time-varying inertia weight.DiscussionSeveral classification and regression problems are utilized to verify the performance of the IBWO-RBF model. In the first stage, the proposed model is applied to UCI dataset classification, nonlinear function approximation, and nonlinear system identification; in the second stage, the model solves the practical problem of power load prediction.ResultsCompared with other existing models, the experiments show that the proposed IBWO-RBF model achieves both accuracy and parsimony in various classification and regression problems.

show abstract

Section: Experiments On Common Classification and Regression Problemsmentioning

confidence: 99%

An RBF neural network based on improved black widow optimization algorithm for classification and regression problems

Liu

Zhou²,

Zhou³

et al. 2023

Front. Neuroinform.

View full text Add to dashboard Cite

show abstract

“…Thus, on the basis of PALM-DE, Meng et al (Meng et al, 2019) introduced a new parameter adaptive DE variant called PaDE to resolve inappropriate adaptation schemes with a novel grouping strategy. In order to provide explicit guidelines for generating appropriate control parameters, Wang et al (Wang et al, 2019) proposed a parameter adaption strategy based on Association Rule Mining methodology, which could extract precise effectual F and CR couple associations during different evolving phases. Sun et al (Sun et al, 2020a) constructed an adaptive parameter settings using decreasing and periodic functions for balancing the global exploration ability and the local exploitation ability.…”

Section: Related Workmentioning

confidence: 99%

Differential evolution with dynamic combination based mutation operator and two-level parameter adaptation strategy

Deng

Lan

et al. 2022

Expert Systems with Applications

View full text Add to dashboard Cite

“…After the first operation, crossover operation is usually given to generate a trial vector U i,G for each couple of donor vector and mutation vector. In the conventional version, the binomial crossover operator is often used as follows [23]:…”

Section: B Crossovermentioning

confidence: 99%

“…Moreover, DE and similar algorithms have been used to optimize ANNs [8], [9], [21]. However, the performances of DE algorithms need more improvements [22], [23].…”

Section: Introductionmentioning

confidence: 99%

A Collective Intelligence Based Differential Evolution Algorithm for Optimizing the Structure and Parameters of a Neural Network

et al. 2020

Self Cite

View full text Add to dashboard Cite

In this paper, a Self-learning Collective Intelligence Differential Evolution (SLCIDE) algorithm was proposed to optimize both the architecture and parameters of a Feedforward Neural Network (FNN). In order to improve the exploration-exploitation capability, a new Collective Intelligence (CI) based mutation operator was proposed by mixing some promising donor vectors in the current population. Besides, a self-learning mechanism which was designed to adaptively select m top ranked donor vectors was developed by using a widely used unsupervised learning method, k-means. As a result, the proposed approach can be more adaptive and statistically powerful on versatile problems. Then, we evaluated the performances of the proposed SLCIDE approach by studying some numerical optimization problems of CEC 2013 with D = 30 and D = 50. Further, the proposed SLCIDE method was applied to train a FNN on four most popular datasets, resulting in very competitive performances. The comprehensive experimental results have demonstrated that the presented SLCIDE method obtain better results compared with other state-of-the-art algorithms. INDEX TERMS Evolutionary artificial neural network, global optimization, collective intelligence (CI), differential evolution (DE).

show abstract

Association rule mining based parameter adaptive strategy for differential evolution algorithms

Cited by 24 publications

References 48 publications

An RBF neural network based on improved black widow optimization algorithm for classification and regression problems

An RBF neural network based on improved black widow optimization algorithm for classification and regression problems

Differential evolution with dynamic combination based mutation operator and two-level parameter adaptation strategy

A Collective Intelligence Based Differential Evolution Algorithm for Optimizing the Structure and Parameters of a Neural Network

Contact Info

Product

Resources

About