Effective Linkage Learning Using Low-Order Statistics and Clustering

Emmendorfer, Leonardo Ramos; Pozo, Aurora

doi:10.1109/tevc.2009.2025455

Cited by 22 publications

(47 citation statements)

References 24 publications

(69 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instances with 2-4 objectives, MOEA/D exhibits the best performance, however it is closely followed by However for more than 9 objectives, random search out-performs the other algorithms. Also, when compared with RM-MEDA, both MACE and MACE-gD perform significantly better for all instances with 2-10 objectives, a fact that supports the theory presented in [36] that EDAs using low order statistics with some form of clustering have potential. Of course, clustering is not used in these versions of the MACE algorithm; this is the subject of future research.…”

Section: Experiments Resultssupporting

confidence: 81%

“…This is fortunate since the CE method is based on sound theoretical principles which can facilitate further analysis of this method. Also, the hypothesis presented in [36], that EDAs based on low order statistics and clustering can be used as an alternative to complex probabilistic models, seems to be supported by the obtained results in Section VII-C. However, as no clustering method is employed in MACE-gD, this does not constitute solid proof but it is certainly a good indication.…”

Section: Discussionmentioning

confidence: 78%

“…To overcome such difficulties researchers proposed ever more elaborate models [34], which of course increase the complexity of the algorithm and in some instances the identification of the optimal model is of comparable complexity to that of the optimization problem necessitating the use of heuristics [35]. Acknowledging this problem has led some researchers to suggest hybridization of EDAs based on simple probabilistic models with some form of clustering [36]. This course is further supported by more recent studies [37].…”

mentioning

confidence: 98%

See 2 more Smart Citations

Generalized decomposition and cross entropy methods for many-objective optimization

Giagkiozis

Purshouse

Fleming

2014

Information Sciences

103

View full text Add to dashboard Cite

Monograph:Giagkiozis, I., Purshouse, R.C. and Fleming, P.J. (2012) ReuseUnless indicated otherwise, fulltext items are protected by copyright with all rights reserved. The copyright exception in section 29 of the Copyright, Designs and Patents Act 1988 allows the making of a single copy solely for the purpose of non-commercial research or private study within the limits of fair dealing. The publisher or other rights-holder may allow further reproduction and re-use of this version -refer to the White Rose Research Online record for this item. Where records identify the publisher as the copyright holder, users can verify any specific terms of use on the publisher's website. TakedownIf you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request. Abstract-Decomposition-based algorithms for multi-objective optimization problems have increased in popularity in the past decade. Although their convergence to the Pareto optimal front (PF) is in several instances superior to that of Pareto-based algorithms, the problem of selecting a way to distribute or guide these solutions in a high-dimensional space has not been explored. In this work, we introduce a novel concept which we call generalized decomposition. Generalized decomposition provides a framework with which the decision maker (DM) can guide the underlying evolutionary algorithm toward specific regions of interest or the entire Pareto front with the desired distribution of Pareto optimal solutions. Additionally, it is shown that generalized decomposition simplifies many-objective problems by unifying the three performance objectives of multi-objective evolutionary algorithms -convergence to the PF, evenly distributed Pareto optimal solutions and coverage of the entire front -to only one, that of convergence. A framework, established on generalized decomposition, and an estimation of distribution algorithm (EDA) based on low-order statistics, namely the cross-entropy method (CE), is created to illustrate the benefits of the proposed concept for many objective problems. This choice of EDA also enables the test of the hypothesis that low-order statistics based EDAs can have comparable performance to more elaborate EDAs.

show abstract

Section: Experiments Resultssupporting

confidence: 81%

Section: Discussionmentioning

confidence: 78%

mentioning

confidence: 98%

See 1 more Smart Citation

Generalized decomposition and cross entropy methods for many-objective optimization

Giagkiozis

Purshouse

Fleming

2014

Information Sciences

103

View full text Add to dashboard Cite

show abstract

“…Estimation of Distribution Algorithms (EDAs) (Larrañaga and Lozano 2002) that employ this technique are (Emmendorfer and Pozo 2009;Pelikan and Goldberg 2000;Tsuji et al 2006;Bosman and Thierens 2002).…”

Section: Clustering In Single Objective Optimizationmentioning

confidence: 99%

An investigation of clustering strategies in many-objective optimization: the I-Multi algorithm as a case study

et al. 2017

View full text Add to dashboard Cite

A variety of general strategies have been applied to enhance the performance of multi-objective optimization algorithms for many-objective optimization problems (those with more than three objectives). One of these strategies is to split the solutions to cover di↵erent regions (clusters) and apply an optimizer to each region with the aim of producing more diverse solutions and achieving a better distributed approximation of the Pareto front. However, the e↵ectiveness of clustering in this context depends on a number of issues, including the characteristics of the objective functions. In this paper we show how the choice of the clustering strategy can greatly influence the behavior an optimizer. We try to find a relation between the characteristics a of multi-objective optimization problem (MOP) and the e ciency of the use of a clustering type in its resolution. Using as a case study, the Iterated Multi-swarm (I-Multi), a recently introduced multi-objective particle swarm optimization (MOPSO) algorithm, we scrutinize the impact that clustering in di↵erent spaces (of variables, objectives, and a combination of both) can have on the approximations of the Pareto front. Furthermore, using two difficult function benchmarks of problems of up to 20 objectives, we evaluate the e↵ect of using di↵erent metrics for determining the similarity between the solutions during the clustering process. Our results confirm the important e↵ect of the clustering strategy on the behavior of multi-objective optimizers. Moreover, we present evidence that some problem characteristics can be used to select the most e↵ective clustering strategy, significantly improving the quality of the Pareto front approximations produced by I-Multi.

show abstract

“…Thus conveying local minimum instead of global minimum. In his work, Emmendorfer [18] showed that a mixture of Gaussians, as a multimodal distribution, is a good alternative to model non-convex search spaces.…”

Section: Pbil and Clusteringmentioning

confidence: 99%

Study on the Use of Evolutionary Techniques for Inference in Gene Regulatory Networks

Palafox

Noman

Iba

2013

Proceedings in Information and Communications Technology

View full text Add to dashboard Cite

Abstract. Inference in Gene Regulatory Networks remains an important problem in Molecular Biology. Many models have been proposed to model the relationships within genes in a DNA chain. Many of these models use Evolutionary Techniques to find the best parameters of specific DNA motifs.In this work, we compare the popular S-System using a powerful evolutionary technique, DPSO, and the novel Recursive Neural Network model, using clustered Population Based Incremental Learning (PBIL). We will use the SOS network for E.coli to do the comparison to finally show how they fare against other techniques in the area of Gene Regulatory Network (GRN) inference.

show abstract

Effective Linkage Learning Using Low-Order Statistics and Clustering

Abstract: Abstract

Cited by 22 publications

References 24 publications

Generalized decomposition and cross entropy methods for many-objective optimization

Generalized decomposition and cross entropy methods for many-objective optimization

An investigation of clustering strategies in many-objective optimization: the I-Multi algorithm as a case study

Study on the Use of Evolutionary Techniques for Inference in Gene Regulatory Networks

Contact Info

Product

Resources

About