稳定性特征选择研究综述

doi:10.3778/j.issn.1002-8331.2406-0410

摘要/Abstract

摘要： 特征选择在高维数据预处理中扮演着重要角色，通过从原始特征集中挑选出最有利于模型性能提升的特征，可以有效地降低数据维数，提高模型的准确性和降低过拟合风险。稳定性是特征选择领域中一个不容忽视的关键研究内容，它指的是特征选择方法对训练样本的微小扰动具有一定的鲁棒性。深入剖析了特征选择过程中产生不稳定性的多重成因；系统归纳并对比了多种提升稳定性的方法，详细阐述了各类方法的目标和评估标准及其独特优势和潜在缺陷；详尽介绍了评估特征选择稳定性的指标的性质，并对稳定性指标进行解析和细致分类；探讨了稳定性特征选择领域存在的问题及对未来的展望，以期为后续的研究和实践提供有价值的参考。

关键词: 特征选择, 稳定性度量, 集成特征选择, 不稳定性

Abstract: Feature selection plays a crucial role in preprocessing high-dimensional data. By selecting the most relevant features from the original set, it effectively reduces data dimensionality, improves model accuracy, and mitigates overfitting risk. Stability is a key focus in feature selection research, referring to the robustness of methods to small perturbations in training samples. Initially, this paper conducts an extensive analysis on various factors contributing to instability during feature selection processes. Subsequently, it systematically summarizes and compares diverse methods aimed at enhancing stability while providing detailed descriptions of their objectives, evaluation criteria as well as unique advantages and potential limitations. Furthermore, it meticulously introduces properties associated with assessing feature selection stability indexes followed by a comprehensive analysis and classification thereof. Ultimately, it delves into discussing challenges within stable feature selections alongside future prospects with an aim to offer valuable insights for subsequent research endeavors.

Key words: feature selection, stability measure, ensemble feature selection, instability

刘梓萱, 杜建强, 罗计根, 黄强, 贺佳, 李益雯, 秦紫瑜. 稳定性特征选择研究综述[J]. 计算机工程与应用, 2025, 61(7): 81-95.

LIU Zixuan, DU Jianqiang, LUO Jigen, HUANG Qiang, HE Jia, LI Yiwen, QIN Ziyu. Review of Stability Feature Selection[J]. Computer Engineering and Applications, 2025, 61(7): 81-95.

参考文献

[1] KALOUSIS A, PRADOS J, HILARIO M. Stability of feature selection algorithms: a study on high-dimensional spaces[J]. Knowledge and Information Systems, 2007, 12(1): 95-116.
[2] KHAIRE U M, DHANALAKSHMI R. Stability of feature selection algorithm: a review[J]. Journal of King Saud University-Computer and Information Sciences, 2022, 34(4): 1060-1073.
[3] NOGUEIRA S, BROWN G. Measuring the stability of feature selection[C]//Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases. Cham: Springer, 2016: 442-457.
[4] YU J M, YIN J J, HUANG S A, et al. A new image super-resolution reconstruction algorithm based on hybrid diffusion model[C]//Proceedings of the 16th Chinese Conference on Image and Graphics Technologies and Applications. Singapore: Springer, 2021: 173-188.
[5] ALELYANI S. Stable bagging feature selection on medical data[J]. Journal of Big Data, 2021, 8(1): 11.
[6] PAU S, PERNICIANO A, PES B, et al. An evaluation of feature selection robustness on class noisy data[J]. Information, 2023, 14(8): 438.
[7] ORRù P, PES B. Feature selection on imbalanced domains: a stability-based analysis[C]//Proceedings of the International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems. Cham: Springer, 2023: 14-27.
[8] HUANG C Y. Feature selection and feature stability measurement method for high-dimensional small sample data based on big data technology[J]. Computational Intelligence and Neuroscience, 2021, 2021(1): 3597051.
[9] WALD R, KHOSHGOFTAAR T M, NAPOLITANO A. Stability of filter- and wrapper-based feature subset selection[C]//Proceedings of the 2013 IEEE 25th International Conference on Tools with Artificial Intelligence. Piscataway: IEEE, 2013: 374-380.
[10] BOMMERT A, RAHNENFüHRER J. Adjusted measures for feature selection stability for data sets with similar features[C]//Proceedings of the 6th International Conference on Machine Learning, Optimization, and Data Science. Cham: Springer, 2020: 203-214.
[11] CHEN X Y, WANG X T, ZHOU J T, et al. Activating more pixels in image super-resolution transformer[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 22367-22377.
[12] SHI J P, LI H, LIU T L, et al. Image super-resolution using efficient striped window transformer[J]. arXiv:2301.09869, 2023.
[13] MIENYE I D, SUN Y X. A survey of ensemble learning: concepts, algorithms, applications, and prospects[J]. IEEE Access, 2022, 10: 99129-99149.
[14] SAEYS Y, ABEEL T, VAN DE PEER Y. Robust feature selection using ensemble feature selection techniques[C]//Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases. Berlin, Heidelberg: Springer, 2008: 313-325.
[15] ABEEL T, HELLEPUTTE T, VAN DE PEER Y, et al. Robust biomarker identification for cancer diagnosis with ensemble feature selection methods[J]. Bioinformatics, 2010, 26(3): 392-398.
[16] GURU D S, SUHIL M, PAVITHRA S K, et al. Ensemble of feature selection methods for text classification: an analytical study[C]//Proceedings of the 17th International Conference on Intelligent Systems Design and Applications, 2018: 337-349.
[17] SALMAN R, ALZAATREH A, SULIEMAN H. The stability of different aggregation techniques in ensemble feature selection[J]. Journal of Big Data, 2022, 9(1): 51.
[18] EFFROSYNIDIS D, ARAMPATZIS A. An evaluation of feature selection methods for environmental data[J]. Ecological Informatics, 2021, 61: 101224.
[19] WANG H Z, HE C Q, LI Z P. A new ensemble feature selection approach based on genetic algorithm[J]. Soft Computing, 2020, 24(20): 15811-15820.
[20] BOLóN-CANEDO V, ALONSO-BETANZOS A. Ensembles for feature selection: a review and future trends[J]. Information Fusion, 2019, 52: 1-12.
[21] PES B, DESSì N, ANGIONI M. Exploiting the ensemble paradigm for stable feature selection: a case study on high-dimensional genomic data[J]. Information Fusion, 2017, 35: 132-147.
[22] MOON M, NAKAI K T. Stable feature selection based on the ensemble L1-norm support vector machine for biomarker discovery[J]. BMC Genomics, 2016, 17(Suppl 13): 1026.
[23] RONDINA J M, HAHN T, DE OLIVEIRA L, et al. SCoRS: a method based on stability for feature selection and mapping in neuroimaging[J]. IEEE Transactions on Medical Imaging, 2014, 33(1): 85-98.
[24] MOROVVAT M, OSAREH A. An ensemble of filters and wrappers for microarray data classification[J]. Mach Learn Appl An Int J, 2016, 3(2): 1-17.
[25] WANG J, XU J, ZHAO C G, et al. An ensemble feature selection method for high-dimensional data based on sort aggregation[J]. Systems Science & Control Engineering, 2019, 7(2): 32-39.
[26] PES B. Ensemble feature selection for high-dimensional data: a stability analysis across multiple domains[J]. Neural Computing and Applications, 2020, 32(10): 5951-5973.
[27] FAHAD A, TARI Z, KHALIL I, et al. An optimal and stable feature selection approach for traffic classification based on multi-criterion fusion[J]. Future Generation Computer Systems, 2014, 36: 156-169.
[28] XIANG F, ZHAO Y L, ZHANG M, et al. Ensemble learning-based stability improvement method for feature selection towards performance prediction[J]. Journal of Manufacturing Systems, 2024, 74: 55-67.
[29] DITTMAN D J, KHOSHGOFTAAR T M, WALD R, et al. Comparing two new gene selection ensemble approaches with the commonly-used approach[C]//Proceedings of the 2012 11th International Conference on Machine Learning and Applications. Piscataway: IEEE, 2012: 184-191.
[30] BOUCHEHAM A, BATOUCHE M. Massively parallel feature selection based on ensemble of filters and multiple robust consensus functions for cancer gene identification[C]//Proceedings of the Science and Information Conference. Cham: Springer, 2015: 93-108.
[31] LIU Y, DIAO X C, CAO J J, et al. Evolutionary algorithms’ feature selection stability improvement system[C]//Proceedings of the 12th International Conference on Bio-inspired Computing: Theories and Applications. Singapore: Springer, 2017: 68-81.
[32] GINSBURG S B, LEE G, ALI S, et al. Feature importance in nonlinear embeddings(FINE): applications in digital pathology[J]. IEEE Transactions on Medical Imaging, 2016, 35(1): 76-88.
[33] LAHMIRI S, SHMUEL A. Detection of Parkinson’s disease based on voice patterns ranking and optimized support vector machine[J]. Biomedical Signal Processing and Control, 2019, 49: 427-433.
[34] LIU Z, WANG R Y, JAPKOWICZ N, et al. Mobile app traffic flow feature extraction and selection for improving classification robustness[J]. Journal of Network and Computer Applications, 2019, 125: 190-208.
[35] LI Y, LI T, LIU H. Recent advances in feature selection and its applications[J]. Knowledge and Information Systems, 2017, 53(3): 551-577.
[36] YU L, DING C, LOSCALZO S, et al. Stable feature selection via dense feature groups[C]//Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2008: 803-811.
[37] LEE I G, YOON S W, WON D. A mixed integer linear programming support vector machine for cost-effective group feature selection: branch-cut-and-price approach[J]. European Journal of Operational Research, 2022, 299(3): 1055-1068.
[38] KAMKAR I, GUPTA S K, PHUNG D, et al. Stable feature selection for clinical prediction: exploiting ICD tree structure using Tree-Lasso[J]. Journal of Biomedical Informatics, 2015, 53: 277-290.
[39] FERNANDEZ-LOZANO C, SEOANE J A, GESTAL M, et al. Texture classification using feature selection and kernel-based techniques[J]. Soft Computing, 2015, 19(9): 2469-2480.
[40] WU T, HAO Y H, YANG B, et al. ECM-EFS: an ensemble feature selection based on enhanced co-association matrix[J]. Pattern Recognition, 2023, 139: 109449.
[41] HU R Y, ZHU X F, CHENG D B, et al. Graph self-representation method for unsupervised feature selection[J]. Neurocomputing, 2017, 220: 130-137.
[42] MOHAMMADI M, SHARIFI NOGHABI H, ABED HODTANI G, et al. Robust and stable gene selection via maximum-minimum correntropy criterion[J]. Genomics, 2016, 107(2/3): 83-87.
[43] 李云. 稳定的特征选择研究[J]. 微型机与应用, 2012, 31(15): 1-2.
LI Y. Research on stable feature selection[J]. Microcomputer & Its Applications, 2012, 31(15): 1-2.
[44] YU L, HAN Y, BERENS M E. Stable gene selection from microarray data via sample weighting[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2012, 9(1): 262-272.
[45] LI K W, YU M X, LIU L, et al. Feature selection method based on weighted mutual information for imbalanced data[J]. International Journal of Software Engineering and Knowledge Engineering, 2018, 28(8): 1177-1194.
[46] LI Y, SI J, ZHOU G J, et al. FREL: a stable feature selection algorithm[J]. IEEE Transactions on Neural Networks and Learning Systems, 2015, 26(7): 1388-1402.
[47] DONG H B, LI T, DING R, et al. A novel hybrid genetic algorithm with granular information for feature selection and optimization[J]. Applied Soft Computing, 2018, 65: 33-46.
[48] KHAIRE U M, DHANALAKSHMI R. Stability investigation of improved whale optimization algorithm in the process of feature selection[J]. IETE Technical Review, 2022, 39(2): 286-300.
[49] SAKAE Y, STRAUB J E, OKAMOTO Y. Enhanced sampling method in molecular simulations using genetic algorithm for biomolecular systems[J]. Journal of Computational Chemistry, 2019, 40(2): 475-481.
[50] ISACHENKO R V, STRIJOV V V. Quadratic programming optimization with feature selection for nonlinear models[J]. Lobachevskii Journal of Mathematics, 2018, 39(9): 1179-1187.
[51] NOGUEIRA S, BROWN G. Measuring the stability of feature selection with applications to ensemble methods[C]//Proceedings of the 12th International Workshop on Multiple Classifier Systems. Cham: Springer, 2015: 135-146.
[52] NOGUEIRA S, SECHIDIS K, BROWN G. On the stability of feature selection algorithms[J]. Journal of Machine Learning Research, 2017, 18(1): 6345-6398.
[53] KUNCHEVA L. A stability index for feature selection[C]//Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, 2007: 421-427.
[54] 李郅琴, 杜建强, 聂斌, 等. 特征选择方法综述[J]. 计算机工程与应用, 2019, 55(24): 10-19.
LI Z Q, DU J Q, NIE B, et al. Summary of feature selection methods[J]. Computer Engineering and Applications, 2019, 55(24): 10-19.
[55] KHOSHGOFTAAR T M, FAZELPOUR A, WANG H J, et al. A survey of stability analysis of feature subset selection techniques[C]//Proceedings of the 2013 IEEE 14th International Conference on Information Reuse & Integration. Piscataway: IEEE, 2013: 424-431.
[56] 刘艺, 曹建军, 刁兴春, 等. 特征选择稳定性研究综述[J]. 软件学报, 2018, 29(9): 2559-2579.
LIU Y, CAO J J, DIAO X C, et al. Survey on stability of feature selection[J]. Journal of Software, 2018, 29(9): 2559-2579.
[57] ABOU-ABBAS L, HENNI K, JEMAL I, et al. Patient-independent epileptic seizure detection by stable feature selection[J]. Expert Systems with Applications, 2023, 232: 120585.
[58] WANG H, KHOSHGOFTAAR T M, WALD R. Measuring robustness of feature selection techniques on software engineering datasets[C]//Proceedings of the 2011 IEEE International Conference on Information Reuse & Integration, 2011: 309-314.
[59] BOMMERT A. Integration of feature selection stability in model fitting[Z]. Dortmund: TU Dortmund University, 2020.
[60] HOSSEINY MARANI A, BAUMER E P S. A review of stability in topic modeling: metrics for assessing and techniques for improving stability[J]. ACM Computing Surveys, 2023, 56(5): 1-32.
[61] MOHANA CHELVAN P, PERUMAL K. A comparative analysis of feature selection stability measures[C]//Proceedings of the 2017 International Conference on Trends in Electronics and Informatics. Piscataway: IEEE, 2017: 124-128.
[62] 江川. 面向高维小样本数据的特征选择稳定性研究[D]. 宜昌: 三峡大学, 2021.
JIANG C. Study on stability of feature selection for high-dimensional small sample data[D]. Yichang: China Three Gorges University, 2021.
[63] STUDENT S, FUJAREWICZ K. Stable feature selection and classification algorithms for multiclass microarray data [J]. Biology Direct, 2012, 7: 1-20.
[64] DERNONCOURT D, HANCZAR B, ZUCKER J D. Analysis of feature selection stability on high dimension and small sample data[J]. Computational Statistics & Data Analysis, 2014, 71: 681-693.
[65] RAKESH D K, ANWIT R, JANA P K. A new ranking-based stability measure for feature selection algorithms[J]. Soft Computing, 2023, 27(9): 5377-5396.
[66] LUSTGARTEN J L, GOPALAKRISHNAN V, VISWESW-
ARAN S. Measuring stability of feature selection in biomedical datasets[C]//AMIA Annual Symposium Proceedings, 2009: 406-410.
[67] AWADA W, KHOSHGOFTAAR T M, DITTMAN D, et al. A review of the stability of feature selection techniques for bioinformatics data[C]//Proceedings of the 2012 IEEE 13th International Conference on Information Reuse & Integration. Piscataway: IEEE, 2012: 356-363.
[68] ZHANG M, ZHANG L, ZOU J F, et al. Evaluating reproducibility of differential expression discoveries in microarray studies by considering correlated molecular changes[J]. Bioinformatics, 2009, 25(13): 1662-1668.
[69] SECHIDIS K, PAPANGELOU K, NOGUEIRA S, et al. On the stability of feature selection in the presence of feature correlations[C]//Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases. Cham: Springer, 2020: 327-342.
[70] ALAIZ-RODRíGUEZ R, PARNELL A C. An information theoretic approach to quantify the stability of feature selection and ranking algorithms[J]. Knowledge-Based Systems, 2020, 195: 105745.
[71] CHELVAN P M, PERUMAL K. A study on selection stability measures for various feature selection algorithms[C]//Proceedings of the 2016 IEEE International Conference on Computational Intelligence and Computing Research. Piscataway: IEEE, 2016: 1-4.
[72] HAMER V, DUPONT P. An importance weighted feature selection stability measure[J]. Journal of Machine Learning Research, 2021, 22(1): 5153-5209.
[73] BARBIERI M C, GRISCI B I, DORN M. Analysis and comparison of feature selection methods towards performance and stability[J]. Expert Systems with Applications, 2024, 249: 123667.
[74] CHEN P T, LI F, WU C W. Research on intrusion detection method based on Pearson correlation coefficient feature selection algorithm[J]. Journal of Physics: Conference Series, 2021, 1757(1): 012054.