基于实例学习和协同子集搜索的特征选择方法

doi:10.11959/j.issn.1000-0801.2017122

Abstract

Abstract:

Feature subset selection is a key problem in such data mining classification tasks.In practice,the filter methods ignore the correlations between genes which are prevalent in gene expression data,additionally,existing methods are not specially conceived to handle the small sample size of the data which is one of the main causes of feature selection instability.In order to deal with these issues,a new hybrid,filter wrapper was proposed,and a cooperative subset search(CSS),was then researched with a classifier algorithm to represent an evaluation system of wrappers.The method was experimentally tested and compared with state-of-the-art algorithms based on several high-dimension allow sample size cancer data sets.Results show that the proposed approach outperforms other methods in terms of accuracy and stability of the selected subset.

Key words: feature selection, hybrid, small sample, classification, stability

CLC Number:

TP312

Xiaoyuan XU,Li HUANG. A feature selection method based on instance learning and cooperative subset search[J]. Telecommunications Science, 2017, 33(6): 105-113.

Figures/Tables 5

References 21

[1]	刘德山, 范雅惠, 闫德勤 ,等. 一种新的去无关基因肿瘤样本分类方法[J]. 辽宁师范大学学报(自然科学版), 2015,38(1): 41-46.
	LIU D S , FAN Y H , YAN D Q ,et al. A novel removing irrelevant gene classification algorithm for tumor samples[J]. Journal of Liaoning Normal University(Natural Science Edition), 2015,38(1): 41-46.
[2]	KOHANE I S , KHO A T , BUTTE A J . Micro arrays for an integrative genomics[M]. Cambridge: MIT PressPress, 2003.
[3]	TOLOSI L , LENGAUER T . Classification with correlated features:unreliability of feature ranking and solutions[J]. Bioinformatics, 2011,27(14): 1986-1994.
[4]	陈骥思, 余艳梅, 殷宇 ,等. 自适应快速FCM彩色图像分割研究[J]. 计算机工程与应用, 2010,46(7): 178-180.
	CHEN J S , YU Y M , YIN Y ,et al. Research on adaptive fast FCM color image segmentation[J]. Computer Engineering and Application, 2010,46(7): 178-180.
[5]	PENG H , LONG F , DING C . Feature selection based on mutual information:criteria of max-dependency,max-relevance,and min-redundancy[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005(27): 1226-1238.
[6]	KOHAVI R , . A study of cross-validation and boots trap for accuracy estimation and model selection[C]// The 14th International Joint Conference on Artificial Intelligence,August 20-25,1995,Montréal,Canada. Burlington:Morgan Kaufmann Publishers, 1995: 1137-1143.
[7]	HUANG D , LAI J H , WANG C D . Combining multiple clustering via crowd agreement estimation and multi-granularity link analysis[J]. Neuro Computing, 2015(170): 240-250.
[8]	ALOM T N , SUDIPTA R . Identification of WBC based on dynamic clustering using modified FCM algorithm with an approach to optimal result[C]// ACM 6th International Conference on Computer and Communication Technology(ICCCT 2015),September 25-27,2015,Allahabad,India. New York:ACM Press, 2015: 461-464.
[9]	潘晓花, 孙文杰, 韦志辉 ,等. 脑MR图像互信息最大的凸优化分割模型[J]. 计算机辅助设计与图形学学报, 2012,24(8): 1082-1089.
	PAN X H , SUN W J , WEI Z H ,et al. The convex optimization segmentation model of MR brain image with maximum mutual information[J]. Journal of Computer Aided Design, 2012,24(8): 1082-1089.
[10]	SHIPP M A , ROSS K N , TAMAYO P ,et al. Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning[J]. Nat.Med, 2002(9): 68-74.
[11]	DYRSKJOT L , THYKJAER T , KRUHOFFER M ,et al. Identifying distinct classes of bladder carcinoma using micro arrays[J]. Nature Genetics, 2003(33): 90-96.
[12]	TROYANSKAYA O , CANTOR M , SHERLOCK G ,et al. Missing value estimation methods for dna microarrays[J]. Bioinformatics, 2001,17(6): 520-525.
[13]	SINGH D,G , FEBBO P , ROSS K ,et al. Gene expression correlates of clinical prostate cancer behavior[J]. Cancer Cell, 2002,1(2): 203-209.
[14]	VAN’ T , VEER L J , DAI H , VANDEVIJVER M J ,et al. Gene expression profiling predicts clinical outcome of breast can cer[J]. Nature, 2002,415(6871): 530-536.
[15]	POMEROY S L , TAMAYO P , GAASENBEEK M ,et al. Prediction of central nervous system embryonal tumour outcome based on gene expression[J]. Nature, 2002,415(6870): 436-442.
[16]	GORDON G , JENSEN R , HSIAO L ,et al. Translation of micro array data into clinically relevant cancer diagnostic tests using gene expression ratios in lung cancer and mesothelioma[J]. Cancer Res, 2002(62): 4963-4967.
[17]	GUYON I , BENHUR A , GUNN S ,et al. Result analysis of the nips 2003 feature selection challenge[C]// International Conference on Neural Information Processing Systems,November 22-25,2004,Vancouver,British Columbia,Canada. New York:ACM Press, 2004: 545-552.
[18]	李鸿健, 代宇, 刘锐 ,等. 云数据中心高能效的虚拟机迁移整合算法研究[J]. 电信科学, 2015,31(1): 71-77.
	LI H J , DAI Y , LIU R ,et al. Energy-efficient virtual machine migration and consolidation algorithm in cloud data center[J]. Telecommunications Science, 2015,31(1): 71-77.
[19]	张建敏, 谢伟良, 杨峰义 ,等. 移动边缘计算技术及其本地分流方案[J]. 电信科学, 2016,32(7): 132-139.
	ZHANG J M , XIE W L , YANG F Y ,et al. Mobile edge computing and application in traffic offloading[J]. Telecommunications Science, 2016,32(7): 132-139.
[20]	李丹, 刘方明, 郭得科 ,等. 软件定义的云数据中心网络基础理论与关键技术[J]. 电信科学, 2014,30(6): 48-59.
	LI D , LIU F M , GUO D K ,et al. Fundamental theory and key technology of software defined cloud data center network[J]. Telecommunications Science, 2014,30(6): 48-59.
[21]	KALOUSIS A , PRADOS J , SANCHEZ J C ,et al. Distilling classification models from cross validation runs:an application to mass spectrometry[C]// The 2004 IEEE International Conference on Tools with Artificial Intelligence,November 15-17,2004,Boca Raton,Florida,USA. New Jersey:IEEE Press, 2004: 113-119.

Metrics

Recommended 0

No Suggested Reading articles found!

样本	特征值1	特征值2	…	特征值d
样本1	w_1,1	w_1,2	…	w_1,d
样本2	w_2,1	w_2,2	…	w_2,d
样本3	w_3,1	w_3,2	…	w_3,d
…	…	…	…	…
样本m	w_m,1	w_m,2	…	w_m,d

数据集	协同子集搜索（SVM算法）			协同子集搜索（KNN算法）
数据集	候选特征子集	终选特征子集	最小分类误差	候选特征子集	终选特征子集	最小分类误差
DLBCL	10	68	0.012 5	5	28	0.050 0
膀胱癌	9	50	0.033 3	9	17	0.033 3
淋巴癌	7	28	0.000 0	5	7	0.000 0
前列腺癌	9	69	0.050 0	2	25	0.099 1
胸膜癌	10	92	0.268 3	9	100	0.288 9
CNS	8	36	0.209 5	1	1	0.231 9
肺癌	5	25	0.005 3	5	129	0.000 0
UCI	14	200	0.063 7	5	189	0.090 8

数据集	DLBCL	膀胱癌	淋巴癌	前列腺癌	胸腺癌	CNS	肺癌	UCI
样本数/个	77	31	45	102	97	60	181	7 000
特征数/个	7 028	3 035	4 025	12 600	24 482	7 129	12 533	5 000

数据集	HIB-CSS （SVM）	HIB-CSS （KNN）	Hyb-Seq	随机化算法	SVM-RFE	FGM	t检验	Relief	mRMR
DLBCL	0.012 5	0.050 0	0.103 9	0.051 9	0.051 9	0.064 9	0.1429	0.064 9	0.077 9
	0.718 5	0.689 3	0.207 8	0.070 9	0.510 2	0.632 6	0.879 1	0.633 5	0.565 7
膀胱癌	0.033 3	0.033 3	0.322 3	0.064 4	0.032 3	0.096 8	0.032 3	0.064 5	0.064 5
	0.544 5	0.592 5	0.738 9	0.066 7	0.437 2	0.499 2	0.687 7	0.465 0	0.403 9
淋巴癌	0.000 0	0.000 0	0.000 0	0.066 7	0.022 2	0.155 6	0.022 2	0.022 2	0.000 0
	0.512 5	0.233 5	0.715 8	0.068 7	0.435 0	0.730 4	0.723 4	0.429 1	0.595 1
前列腺癌	0.050 0	0.099 1	0.068 5	0.088 2	0.127 5	0.480 4	0.107 8	0.215 7	0.078 4
	0.627 8	0.655 4	0.342 3	0.021 6	0.586 3	0.911 1	0.794 4	0.627 5	0.709 6
胸膜癌	0.268 2	0.288 9	0.525 9	0.268 0	0.299 0	0.505 2	0.525 8	0.443 3	0.278 4
	0.488 4	0.471 8	1.000 0	0.016 8	0.380 6	1.000 0	1.000 0	0.362 8	0.335 5
CNS	0.209 5	0.231 9	0.316 7	0.333 3	0.300 0	0.500 0	0.366 7	0.366 7	0.300 0
	0.208 6	0.622 2	0.121 7	0.044 7	0.491 1	0.964 4	0.397 0	0.497 4	0.379 2
肺癌	0.005 3	0.000 0	0.027 5	0.055 2	0.005 5	0.071 8	0.016 6	0.011 0	0.005 5
	0.495 8	0.880 4	0.121 7	0.011 2	0.483 1	1.000 0	0.836 8	0.840 9	0.661 0
UCI	0.063 8	0.090 8	0.096 7	0.133 0	0.079 9	0.047 6	0.096 3	0.471 1	0.045 8
	0.233 3	0.101 0	0.156 7	0.020 2	0.403 2	0.506 3	0.767 1	0.589 1	0.951 2

A feature selection method based on instance learning and cooperative subset search

RichHTML

PDF下载

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 5

References 21

Related Articles 15

Metrics

Recommended 0

[1]	Lu HAN, Weiyu CHEN, Fei ZHANG, Jianfeng HE, Huaizhen SU. Big data classification method of non relational distributed submission information under differentiated requirements [J]. Telecommunications Science, 2023, 39(6): 114-121.
[2]	Hui WANG, Qixin TAI, Liu LIU, Hong WANG, Rongfang SONG. Low-power transmission method for uplink millimeter-wave massive MIMO-NOMA system based on GSIC [J]. Telecommunications Science, 2023, 39(1): 51-59.
[3]	Yan GAO, Jian SHI, Shengyu MA, Bolin MA, Guangxue YUE. DropBlock based bimodal hybrid neural network for wireless communication modulation recognition [J]. Telecommunications Science, 2022, 38(5): 75-86.
[4]	Da WANG, Tao SUN, Xiaowen SUN, Jianchao GUO, Jie SUN, Guo CHEN, Aidong YANG, Ye OUYANG. Study on digital twins in network lifecycle management [J]. Telecommunications Science, 2022, 38(4): 138-145.
[5]	Yong SONG, Zhiwei YAN, Yukun QIN, Dongming ZHAO, Xiaozhou YE, Yuanyuan CHAI, Ye OUYANG. Customer service complaint work order classification based on matrix factorization and attention multi-task learning [J]. Telecommunications Science, 2022, 38(2): 103-110.
[6]	Zhenhua ZHANG, Siyue SUN, Gaosai LIU, Long WANG, Xinglong JIANG, Lin DONG, Guang LIANG. A survey of optical/electric hybrid switching technology for satellite Internet [J]. Telecommunications Science, 2022, 38(11): 1-10.
[7]	Xiaojian ZHANG, Jiaxuan FEI, Haitao JIANG, Qigui YAO. Security risk analysis of power 5G hybrid networking [J]. Telecommunications Science, 2022, 38(1): 132-139.
[8]	Xie LU, Lei XU, Manjun ZHANG. Safe hierarchical virtual network mapping method based on clustering [J]. Telecommunications Science, 2021, 37(9): 112-117.
[9]	Jiyi WU, Wenjuan LI, Jian CAO, Shiyou QIAN, Qifei ZHANG, Rajkumar BUYYA. AIoT: a taxonomy, review and future directions [J]. Telecommunications Science, 2021, 37(8): 1-17.
[10]	Shuai KANG, Jianwu ZHANG, Zunjie ZHU, Guofeng TONG. An improved YOLOv4 algorithm for pedestrian detection in complex visual scenes [J]. Telecommunications Science, 2021, 37(8): 46-56.
[11]	Haoyi CHEN, Guangqiu LI, Hui LI. Hybrid precoding method for millimeter-wave massive MIMO systems based on IAFS algorithm [J]. Telecommunications Science, 2021, 37(8): 77-84.
[12]	Jin ZHAO, Xiaojun YANG. Application on text classification of telecom user complaints based on GRW and FastText model [J]. Telecommunications Science, 2021, 37(6): 125-131.
[13]	Guoqing LIU, Xingqi WANG, Dan WEI, Jinglong FANG, Yanli SHAO. Feature selection method for software defect number prediction based on maximum information coefficient [J]. Telecommunications Science, 2021, 37(5): 133-147.
[14]	Yue GU, Dan LI, Kaihui GAO. Research on network traffic classification based on machine learning and deep learning [J]. Telecommunications Science, 2021, 37(3): 105-113.
[15]	Dongfeng MAO, Man JIA, Xiaoming HE, Zhihua LIU. Network telemetry technology and its application in automatic network operation and maintenance [J]. Telecommunications Science, 2021, 37(2): 154-163.