检索结果-内蒙古大学图书馆

Informatica (Slovenia) 2024年第1期48卷 57-68页

作者： Hamla, Hind Ghanem, Khadoudja Laboratory of Modelling and Implementation of Complex System Department of Computer Science University of Abdelhamid Mehri Constantine 2 Constantine Algeria

Microarray data analysis has played a significant role in disease diagnosis and tumor type identification over the last two decades. However, due to the curse of dimensionality issues, microarray data classification remains a challenging task. This issue arises from a situation where the number of features is large, but the number of samples is small. As a result, dimension reduction techniques, specifically feature selection methods, are critical for removing non-informative features and improving cancer classification. This paper presents a Filter-embedded hybrid feature selection method to address the gene selection challenge in microarray data analysis. First, it selects the features with the highest Fisher score to create a candidate subset for the next embedded stage. Second, the proposed method employs support vector machine-recursive feature elimination (SVM-RFE) on the candidate subset to identify the optimal set of features to enhance cancer classification. Extensive experiments were conducted with ten high-dimensional microarray datasets to assess the efficacy of the proposed approach. The results show that the proposed method improves classifier performance significantly regarding classification accuracy, number of selected features, and computational efficiency. © 2024 Slovene Society Informatika. All rights reserved.

关键词： Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

A Comparative Study of Filter Feature Selection Methods on Microarray Data 1

引用

12th International Conference on Information systems and Advanced Technologies, ICISAT 2022

作者： Hamla, Hind Ghanem, Khadoudja Laboratory of Modelling and Implementation of Complex System University of Abdelhamid Mehri Constantine 2 El Khroub Algeria

ISBN: (数字)9783031253447

ISBN: (纸本)9783031253430

Microarray data technology offers a great opportunity for the early diagnosis of diseases. However, the classification of microarray data can be difficult for machine learning classifiers because of the large number of features associated with a small number of samples. Feature selection is an efficient reduction method that eliminates non-informative features and preserves relevant features to improve the performance of classifiers. Filter feature selection methods have very low computational time thus, they are well suited for microarray data. In this paper, we review several common filter methods in order to evaluate their effect on six well-known classifiers (random forest, logistic regression, KNN, decision tree, and SVM). Experiments were conduct on ten microarray datasets, the results indicate that using univariate filter feature selection methods (MI, F-score, Fisher score) outperformed multivariate filter feature selection methods (MRMR, CMIM). © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Feature selection Filter methods Machine learning Multivariate methods Univariate methods

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：