Feed-Forward Neural Network Training by Hybrid Bat Algorithm

Milosevic, Stefan; Bezdan, Timea; Živković, Miodrag; Bačanin, Nebojša; Strumberger, Ivana; Tuba, Milan

doi:10.1007/978-3-030-68527-0_4

Cited by 44 publications

(13 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We have empirically determined best values for mpd and mpi, as shown in Eqs. ( 26) and (27), respectively.…”

Section: Working Details Of Proposed Goqrfamentioning

confidence: 99%

“…On the other hand, swarm intelligence algorithms are inspired by social behavior, collaboration, and communication among groups of relatively primitive individuals such as bees, ants, bats, fish, and fireflies. Swarm intelligence approaches were utilized with great success in solving a broad range of real-world problems in practice, including prolonging the network lifetime, clustering, and sensor localization in wireless sensor networks [4,48,50], cloud optimization problems [2,8,10], artificial neural networks optimization [7,20,27,36], machine learningbased COVID-19 cases prediction [49,52], and MRI classification optimization to name a few [5,9].…”

Section: Nature Inspired and Swarm Intelligence Metaheuristicsmentioning

confidence: 99%

See 1 more Smart Citation

Modified firefly algorithm for workflow scheduling in cloud-edge environment

Bačanin

Živković

Bezdan

et al. 2022

Neural Comput & Applic

Self Cite

105

View full text Add to dashboard Cite

Edge computing is a novel technology, which is closely related to the concept of Internet of Things. This technology brings computing resources closer to the location where they are consumed by end-users—to the edge of the cloud. In this way, response time is shortened and lower network bandwidth is utilized. Workflow scheduling must be addressed to accomplish these goals. In this paper, we propose an enhanced firefly algorithm adapted for tackling workflow scheduling challenges in a cloud-edge environment. Our proposed approach overcomes observed deficiencies of original firefly metaheuristics by incorporating genetic operators and quasi-reflection-based learning procedure. First, we have validated the proposed improved algorithm on 10 modern standard benchmark instances and compared its performance with original and other improved state-of-the-art metaheuristics. Secondly, we have performed simulations for a workflow scheduling problem with two objectives—cost and makespan. We performed comparative analysis with other state-of-the-art approaches that were tested under the same experimental conditions. Algorithm proposed in this paper exhibits significant enhancements over the original firefly algorithm and other outstanding metaheuristics in terms of convergence speed and results’ quality. Based on the output of conducted simulations, the proposed improved firefly algorithm obtains prominent results and managed to establish improvement in solving workflow scheduling in cloud-edge by reducing makespan and cost compared to other approaches.

show abstract

“…We have empirically determined best values for mpd and mpi, as shown in Eqs. ( 26) and (27), respectively.…”

Section: Working Details Of Proposed Goqrfamentioning

confidence: 99%

Section: Nature Inspired and Swarm Intelligence Metaheuristicsmentioning

confidence: 99%

Modified firefly algorithm for workflow scheduling in cloud-edge environment

Bačanin

Živković

Bezdan

et al. 2022

Neural Comput & Applic

Self Cite

105

View full text Add to dashboard Cite

show abstract

“…Recently, one of the most important fields of interest has been the hybrid approach with SI and machine learning. The number of publications in this domain increased drastically in recent years; some of the most prominent works include hyperparameter optimization [3,18,19], feature selection problems [2], time series prediction tasks, e.g., estimation of COVID-19 cases [6,20], and neural network training [21,22].…”

Section: Preliminaries and Related Workmentioning

confidence: 99%

Performance of a Novel Chaotic Firefly Algorithm with Enhanced Exploration for Tackling Global Optimization Problems: Application for Dropout Regularization

et al. 2021

Self Cite

View full text Add to dashboard Cite

Swarm intelligence techniques have been created to respond to theoretical and practical global optimization problems. This paper puts forward an enhanced version of the firefly algorithm that corrects the acknowledged drawbacks of the original method, by an explicit exploration mechanism and a chaotic local search strategy. The resulting augmented approach was theoretically tested on two sets of bound-constrained benchmark functions from the CEC suites and practically validated for automatically selecting the optimal dropout rate for the regularization of deep neural networks. Despite their successful applications in a wide spectrum of different fields, one important problem that deep learning algorithms face is overfitting. The traditional way of preventing overfitting is to apply regularization; the first option in this sense is the choice of an adequate value for the dropout parameter. In order to demonstrate its ability in finding an optimal dropout rate, the boosted version of the firefly algorithm has been validated for the deep learning subfield of convolutional neural networks, with respect to five standard benchmark datasets for image processing: MNIST, Fashion-MNIST, Semeion, USPS and CIFAR-10. The performance of the proposed approach in both types of experiments was compared with other recent state-of-the-art methods. To prove that there are significant improvements in results, statistical tests were conducted. Based on the experimental data, it can be concluded that the proposed algorithm clearly outperforms other approaches.

show abstract

“…SI was used to optimize the task scheduling by minimizing the overall time required to execute all tasks and the overall cost [20] [21] [22]. In the domain of machine learning, SI was utilized to design the convolutional neural networks [23], feed-forward neural network training [24] etc. Nature inspired metaheuristics also show very promising results in time-series prediction, the feature that was used to predict the Covid-19 cases [6] [25].…”

Section: Survey Of Nature-inspired Approachesmentioning

confidence: 99%

Nature-Inspired Approaches in Software Testing Optimization

Živković¹,

Živković²,

Džakula³

et al. 2021

Proceedings of the International Scientific Conference - Sinteza 2021

View full text Add to dashboard Cite

Software development is considered to be a fast growing industry that drives numerous modern world domains forward. At the same time, software testing and quality assurance, that is equally important branch of software industry, resides in the shadows, far away from the spotlights. The goal of the software testing is to detect defects in the software and ensure that it has sufficient quality prior to the release.The main objective of the software testing can be defined as finding the minimal test suite that is still efficient enough to maintain the certain software quality. Since the process of test cases generation performs a search for an optimal test suite in a huge search space, and keeping in mind that the swarm intelligence metaheuristics have already proven to be efficient optimizers in other domains, it makes sense to apply swarm intelligence algorithms to the process of test cases generation as well. By utilizing this approach, it could be also possible to reduce the cost required for testing. This paper provides a survey of recent applications of swarm intelligence algorithms in the domain of software testing.

show abstract

Feed-Forward Neural Network Training by Hybrid Bat Algorithm

Cited by 44 publications

References 29 publications

Modified firefly algorithm for workflow scheduling in cloud-edge environment

Modified firefly algorithm for workflow scheduling in cloud-edge environment

Performance of a Novel Chaotic Firefly Algorithm with Enhanced Exploration for Tackling Global Optimization Problems: Application for Dropout Regularization

Nature-Inspired Approaches in Software Testing Optimization

Contact Info

Product

Resources

About