Multi-objective path relinking for bi-clustering: Application to microarray data

Abstract : In this work we deal with a multiobjective biclustering problem applied to microarray data. MOBI nsga [21] is one of the multiobjective metaheuristics that have been proposed to solve a new multiobjective formulation of the biclustering problem. Using MOBI nsga , biclusters of good quality can be extracted. However, the generated front approximation contains a lot of gaps. Using path relinking strategies, our aim is to improve the generated front’s quality by filling the gaps with new solutions. Therefore, we propose a general scheme PR-MOBI nsga of different possible hybridization of MOBI nsga with path relinking strategies. A comparison of different PR-MOBI nsga hybridizations is performed. Experimental results on reel data sets show that PR-MOBI nsga allows to extract new interesting solutions and to improve the Pareto front approximation generated by MOBI nsga .
Document type :
Conference papers
Liste complète des métadonnées

https://hal.inria.fr/hal-01107757
Contributor : Talbi El-Ghazali <>
Submitted on : Wednesday, January 21, 2015 - 3:06:18 PM
Last modification on : Friday, March 22, 2019 - 1:36:16 AM

Links full text

Identifiers

Citation

Khedidja Seridi, Laetitia Jourdan, El-Ghazali Talbi. Multi-objective path relinking for bi-clustering: Application to microarray data. EMO 2013 : 7th Int. Conf. On Evolutionary Multi-criterion Optimization, Mar 2014, Sheffield, United Kingdom. pp.200-214, ⟨10.1007/978-3-642-37140-0_18⟩. ⟨hal-01107757⟩

Share

Metrics

Record views

212