Efficient high-dimension feature selection based on enhanced equilibrium optimizer; Expert Systems with Applications; Vol. 187
| Parent link: | Expert Systems with Applications Vol. 187.— 2022.— [115882, 23 p.] |
|---|---|
| Hoofdauteur: | |
| Coauteur: | |
| Andere auteurs: | |
| Samenvatting: | Title screen Feature selection (FS) is an important task in any classification process and aims to choose the smallest features number that yields higher classification accuracy. FS can be formulated as a combinatorial NP-hard problem for which robust metaheuristics are used as efficient wrapper-based FS approaches. However, when applied for high dimensional datasets that present large features number and few samples, the effectiveness of such wrapper-metaheuristics degraded, and their computation costs increased. To tackle this problem, we propose in this paper a hybrid FS approach based on the ReliefF filter method and a novel metaheuristic Equilibrium Optimizer (EO). The proposed method, called RBEO-LS, is composed of two phases. In the first phase, the ReliefF algorithm is used as a preprocessing step to assign weights for features, which estimate their relevance to the classification task. In the second phase, the binary EO (BEO) is used as a wrapper search approach. The features are ranked according to their weights and are used for the initialization of the BEO population. We embedded the BEO with a local search strategy to improve its performance by adding relevant features and removing redundant ones from the features subset guided by the features ranking and the Pearson coefficient correlation. The performance of the developed algorithm has been evaluated on sixteen UCI datasets and ten high dimensional biological datasets. The UCI datasets contain a high number of samples and a small or medium number of features. The biological datasets present a high number of features with few samples. The results demonstrate that the use of the ReliefF algorithm and the local search strategy improves the performance of the EO algorithm. The results also show the superiority of the RBEO-LS, among other state-of-the-art approaches. |
| Taal: | Engels |
| Gepubliceerd in: |
2022
|
| Onderwerpen: | |
| Online toegang: | https://doi.org/10.1016/j.eswa.2021.115882 |
| Formaat: | Elektronisch Hoofdstuk |
| KOHA link: | https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=668722 |
MARC
| LEADER | 00000naa0a2200000 4500 | ||
|---|---|---|---|
| 001 | 668722 | ||
| 005 | 20260807071633.0 | ||
| 035 | |a (RuTPU)RU\TPU\network\39959 | ||
| 035 | |a RU\TPU\network\37208 | ||
| 090 | |a 668722 | ||
| 100 | |a 20230119d2022 k||y0rusy50 ba | ||
| 101 | 0 | |a eng | |
| 102 | |a GB | ||
| 135 | |a drcn ---uucaa | ||
| 181 | 0 | |a i | |
| 182 | 0 | |a b | |
| 200 | 1 | |a Efficient high-dimension feature selection based on enhanced equilibrium optimizer |f S. Ouadfel, A. M. Mokhamed Elsaed (Mohamed Abd Elaziz) | |
| 203 | |a Text |c electronic | ||
| 300 | |a Title screen | ||
| 330 | |a Feature selection (FS) is an important task in any classification process and aims to choose the smallest features number that yields higher classification accuracy. FS can be formulated as a combinatorial NP-hard problem for which robust metaheuristics are used as efficient wrapper-based FS approaches. However, when applied for high dimensional datasets that present large features number and few samples, the effectiveness of such wrapper-metaheuristics degraded, and their computation costs increased. To tackle this problem, we propose in this paper a hybrid FS approach based on the ReliefF filter method and a novel metaheuristic Equilibrium Optimizer (EO). The proposed method, called RBEO-LS, is composed of two phases. In the first phase, the ReliefF algorithm is used as a preprocessing step to assign weights for features, which estimate their relevance to the classification task. In the second phase, the binary EO (BEO) is used as a wrapper search approach. The features are ranked according to their weights and are used for the initialization of the BEO population. We embedded the BEO with a local search strategy to improve its performance by adding relevant features and removing redundant ones from the features subset guided by the features ranking and the Pearson coefficient correlation. The performance of the developed algorithm has been evaluated on sixteen UCI datasets and ten high dimensional biological datasets. The UCI datasets contain a high number of samples and a small or medium number of features. The biological datasets present a high number of features with few samples. The results demonstrate that the use of the ReliefF algorithm and the local search strategy improves the performance of the EO algorithm. The results also show the superiority of the RBEO-LS, among other state-of-the-art approaches. | ||
| 461 | |t Expert Systems with Applications | ||
| 463 | |t Vol. 187 |v [115882, 23 p.] |d 2022 | ||
| 610 | 1 | |a электронный ресурс | |
| 610 | 1 | |a труды учёных ТПУ | |
| 610 | 1 | |a feature selection | |
| 610 | 1 | |a equilibrium optimizer | |
| 610 | 1 | |a high-dimension data | |
| 610 | 1 | |a relief | |
| 610 | 1 | |a local search strategy | |
| 610 | 1 | |a оптимизаторы | |
| 610 | 1 | |a равновесие | |
| 610 | 1 | |a рельеф | |
| 610 | 1 | |a локальный поиск | |
| 700 | 1 | |a Ouadfel |b S. |g Salima | |
| 701 | 1 | |a Mokhamed Elsaed (Mohamed Abd Elaziz) |b A. M. |c Specialist in the field of informatics and computer technology |c Professor of Tomsk Polytechnic University |f 1987- |g Akhmed Mokhamed |3 (RuTPU)RU\TPU\pers\46943 | |
| 712 | 0 | 2 | |a Национальный исследовательский Томский политехнический университет |b Инженерная школа информационных технологий и робототехники |c 2017- |x TPU |7 ca |8 rus |9 28330 |
| 801 | 0 | |a RU |b 63413507 |c 20230119 |g RCR | |
| 856 | 4 | |u https://doi.org/10.1016/j.eswa.2021.115882 | |
| 942 | |c CF | ||