.png){kind=icon}
.png){kind=icon}
Return
A gene selection algorithm using simplified swarm optimization with multi-filter ensemble technique
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2020.106994.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Gene expression data has been successfully applied to various purposes, especially for cancer classification. The challenges confronted by the development of effective classifiers for expression data are the curse of dimensionality and over-fitting. Gene selection is an efficient and effective method for overcoming the above difficulties and enhancing the predictive accuracy of a classifier. This work presents a hybrid gene selection method based on a novel multi-filter ensemble technique and simplified swarm optimization (SSO). This proposed method was comprised of two phases. In the first phase, the ensemble technique was developed via VIKOR to integrate the outcomes of multiple filters to preselect the most informative gene subset from the original dataset. In the second phase, SSO was tailored with a new gene pruning strategy, re-initialization scheme, and support vector machine as a wrapper to further search the optimal gene subset on the space of the candidate genes selected in the first phase. As evidence of the performance of the proposed method, extensive experiments on nineteen microarray datasets were carried out. The computational results compared favorably with well-known methods in the literature. Statistical analysis indicated that the proposed method was highly competitive and could be an effective tool for microarray analysis. (c) 2020 Elsevier B.V. All rights reserved.
Keywords:
Gene selection
Simplified swarm optimization
Multi-filter ensemble technique
MRMR
Support vector machine
Classification
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
6.6
Papers:
1.4W
Citations:
4.8W