摘要
随着大数据时代的到来,数据的规模和特征维度呈现爆炸式增长,这给数据处理带来了前所未有的挑战。特征选择作为数据预处理的关键环节,在处理大规模高维数据时显得尤为重要。而进化计算方法因其出色的全局搜索能力和高效的优化性能,越来越多的研究者开始对其进行研究,其在大规模高维特征选择中得到了广泛的应用。本文首先介绍了大规模高维数据处理的重要性;然后简单介绍了部分经典和较新的进化计算方法,并详细介绍了其在大规模高维特征选择中的应用情况;最后对目前进化计算在大规模高维特征选择中存在的问题进行总结,并展望了其未来的发展方向。
With the advent of the big data era,the scale and feature dimensions of data show explosive growth,which brings unprecedented challenges to data processing.Feature selection,as a key link in data preprocessing,is particularly important when processing large-scale high-dimensional data.Due to its excellent global search capabilities and efficient optimization performance,more and more researchers begin to study the evolutionary computing method,and it is widely used in large-scale high-dimensional feature selection.This paper first introduces the importance of large-scale high-dimensional data processing.Then some classic and newer evolutionary calculation methods are briefly introduced,and their applications in large-scale high-dimensional feature selection are introduced in detail.Finally,the application of evolutionary computing in large-scale high-dimensional feature selection is summarized and its future development direction is prospected.
作者
叶志伟
王巧
周雯
王明威
蔡婷
何其祎
YE Zhiwei;WANG Qiao;ZHOU Wen;WANG Mingwei;CAI Ting;HE Qiyi(School of Computer Science,Hubei University of Technology,Wuhan 430068,China)
出处
《北方工业大学学报》
2024年第2期8-19,共12页
Journal of North China University of Technology
基金
基于杂交育种协同进化蚁群算法的工业大数据特征选择研究项目(62376089)
关键词
特征选择
进化计算
全局搜索
数据预处理
机器学习
feature selection
evolutionary computation
global search
data preprocessing
machine learning
