Parallel Implementation of MOEA/D with Parallel Weight Vectors for Feature Selection

In machine learning field, feature selection can be treated as a bi-objective optimization problem. It is reported that a decomposition-based evolutionary multi-objective optimization algorithm (i.e., MOEA/D-STAT) has good diversity performance when coping with feature selection. However, feature se...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics S. 1524 - 1531
Hauptverfasser: Liao, Weiduo, Ishibuchi, Hisao, Meng Pang, Lie, Shang, Ke
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 11.10.2020
Schlagworte:
Computational modeling
Cybernetics
Feature extraction
Feature selection
island models
Machine learning
Machine learning algorithms
Master-slave
master-slave models
MOEA/D-STAT
Optimization
parallel weight vectors
parallelization
ISSN:2577-1655
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In machine learning field, feature selection can be treated as a bi-objective optimization problem. It is reported that a decomposition-based evolutionary multi-objective optimization algorithm (i.e., MOEA/D-STAT) has good diversity performance when coping with feature selection. However, feature selection is also a time-consuming problem considering a large dataset it involves. The computation time can be easily reduced by introducing the parallelization into MOEA/D-STAT, thanks to the decomposition idea of MOEA/D. To the best of our knowledge, this is the first attempt to implement the parallelization of MOEA/D-STAT for feature selection. In this paper, we consider both master-slave models and island models, which are two different approaches of parallelization. In the master-slave models, different offspring assignment mechanisms are considered. In the island models, different island size specification mechanisms are examined. Our experimental results show that the master-slave models can achieve higher speedup and better performance than the island models.
ISSN:2577-1655
DOI:10.1109/SMC42975.2020.9283272