数据挖掘中适用于分类的时序数据特征提取方法
|
摘要
特征提取在提高分类的准确性中起着非常关键的作用.对时序特征提取的方法进行归纳分类,将有利于对特征提取整体性,全面性的认识.回顾现有的时间序列中特征提取的方法,将其总结为四大类,它们分别是基于基本统计方法的特征提取、基于模型的特征提取、基于变换的特征提取、基于分形维数的特征提取.针对每一类的特征提取方法,进一步研究了它相应的分类方法和它在时间序列数据中的应用邻域.
The main contributions of this paper are:1) The main feature extraction approaches are classified into four categories;2) The main idea of each category is analyzed,the advantages and disadvantages are pointed out;3) The guidelines of choosing suitable feature extraction approach is suggested.
The main contributions of this paper are:1) The main feature extraction approaches are classified into four categories;2) The main idea of each category is analyzed,the advantages and disadvantages are pointed out;3) The guidelines of choosing suitable feature extraction approach is suggested.
引文
1陈晓云,吴本昌,韩海涛.基于多维时间序列数据挖掘的降雨天气模型研究.计算机工程与设计,2010,4.
2 Mark AKHB.Last Data Mining In Time Series Databases,World Scientific,2004.
3韦东兴,陈晓云,徐荣聪.基于角点检测的图像形状特征提取方法.计算机工程,2010,4.
4加玉涛,罗志增.肌电信号特征提取方法综述.电子器件,2007:326-330.
5刘慧,谢洪波,卫星.基于模糊熵的脑电睡眠分期特征提取与分类.数据采集与处理,2010,25(4):484-489.
6 Greene BR,Boylan GB,Reilly RB.Connolly Combination ofEEG and ECG for improved automatic neonatal seizuredetection,Clin Neurophysiol 2007,118:1348-1359.
7 Gao GQ.Computerised detection and classification of fivecardiac conditions,AUT University.
8商卫波.心电信号自动分析与诊断处理方法研究.西安:西北工业大学,2005.
9葛丁飞,邵宇权,蒋惠忠.基于双导联Ecg和多变量回归模型的远程心电诊断算法研究.航天医学与医学工程,2004:355-359.
10 G.D.-F.H.B.-P.X.Xin-Jian.Study of feature extraction basedon autoregressive modeling in ECG automatic diagnosis.Acta Automatica Sinica,2007,33(5).
11葛丁飞,夏顺仁.Ar模型在远程心电诊断中的应用.中国生物医学工程学报,2004:222-229.
12钱争鸣.Arch族计量模型在金融市场研究中的应用.厦门大学学报(哲学社会科学版),2000:126-129.
13余素红,张世英.Sv与Garch模型对金融时间序列刻画能力的比较研究.系统工程,2002:28-33.
14 Kalantzis T,Papanastassiou D.Classification of GARCHtime series:an empirical investigation.Applied FinancialEconomics,2008,18:759-764.
15 Mrchen F.Time series feature extraction for data miningusing DWT and DFT,2003.
16 Valenza G,Lanata A,Ferro M,Scilingo EP.Real-timediscrimination of multiple cardiac arrhythmias for wearablesystems based on neural networks.Computers in Cardiology,2008.1053-1056.
17郭躬德,王晖.DavidBell时间序列数据分析与预处理.小型微型计算机系统,2003:2228-2232.
18江近仁,孙景江.以反应谱和功率谱密度函数表征的强震运动的统计特性.中国地震,1994:327-340.
19 Amini AM.Qualitative features extraction from sensor datausing Short-Time Fourier Transform,Visual Information Pr-ocessing XIV 5817(2005)289-295.
20 Holmberg M,Gelbart D,Hemmert W.Automatic speechrecognition with an adaptation model motivated by auditoryprocessing,Audio,Speech,and Language Processing.IEEETrans.on,2006,14:43-49.
21 Xu H,Tan ZH,Dalsgaard P,Lindberg B.Exploitation ofspectral variance to improve robustness in speech recogni-tion.Electronics Letters,2006,42:312-314.
22 F.H.a.H.Ney Noise Level Normalization and ReferenceAdaptation for Robust Speech Recognition.InternationalWorkshop on Automatic Speech Recognition:Challengesfor the New Millennium.2000.
23 Hilger F,Ney H.Quantile based histogram equalization fornoise robust large vocabulary speech recognition,Audio,Speech,and Language Processing.IEEE Trans.on,2006,14:845-854.
24高瑞华.多种预处理方法在语音检测中应用效果的比较研究.控制理论与控制工程[硕士学位论文].浙江工业大学,2004.64.
25 J.R.G.Townshend,C.O.Justice and V.Kalb Characterizationand Classification of South American Land Cover TypesUsing Satellite Data.International Journal of RemoteSensing,1987,8:1189-1207.
26邢玉娟,谭萍,李明.一种新的说话人识别序列特征提取方法.兰州理工大学学报,2009,35(4):98-102.
27 Scholkopf SAB,Muller KR.Nonlinear component analysis asa kernel eigenvalue problem.Neural Computaion.
28杨绍华.一种基于核主成分分析的人脸识别方法.河北科技师范学院学报,2008:45-48,62.
29周小程,马向玲,范洪达,庞文强.基于核函数分类的多维时序特征选择方法应用.电光与控制,2010,17(7):74-77.
30侯振雨,蔡文生,邵学广.主成分分析-支持向量回归建模方法及应用研究.分析化学,2006:617-620.
31楼蒋.Method of Feature Extraction Suitable for Hyperdi-mensional Time Series Data.浙江科技学院学报,2007,19(2).
32胡利平,刘宏伟,吴顺君.基于两级2dpca的Sar目标特征提取与识别.电子与信息学报,2008:1722-1726.
33李志农,曾明如,韩捷,何永勇.主分量分析和因子隐Markov模型在机械故障诊断中的应用.机械强度,2007:25-29.
34朱帮助.基于特征提取的选择性神经网络集成方法.计算机科学,2008:132-133,172.
35刘爱霞,刘正军,王静.基于Pca变换和神经元网络分类方法的中国森林制图研究.长江流域资源与环境,2006:19-24.
36王海燕,刘鲁,刘玲.基于Gra和pca的Bp神经网络应用研究.管理评论,2007:50-54.
37张静远,张冰.基于小波变换的特征提取方法分析.信号处理,2000:156-162,155.
38马陈,马国亮.心电图的小波变换特征提取.中国科技信息,2006(1A).
39 Andreao RV,Dorizzi B,Boudy J.ECG signal analysisthrough hidden Markov models.IEEE Trans.on BiomedicalEngineering,2006,53:1541-1549.
40胡博,卫宏儒,廖福成.小波变换在股票分析中的应用.系统管理学报,2007:365-369.
41金解放,赵奎,王晓军,赵康.岩石声发射信号处理小波基选择的研究.矿业研究与开发,2007:12-15.
42朱江梅,吴兴方.小波包变换压缩与地震数据处理.中国海上油气(工程),2006:169-173.
43田喜英.主元分析和非线性方法在表面肌电信号中的应用研究[硕士学位论文].上海交通大学.2009.
44 Gupta AMR,Singh K.A Time-Series-Based FeatureExtraction Approach for Prediction of Protein StructuralClass,EURASIP.Journal on Bioinformatics and SystemsBiology,2008.
45 Palancz B,Benyo B.Classification of time series viawavelet subband analysis using support vector machineclassifier.Periodica Polytechnica Electrical Engineering 50,2006:129-140.
46郑继明,邢峰,吴渝,李婧.基于小波变换和支持向量机的音频分类.重庆邮电大学学报(自然科学版),2008:212-216.
47王春.基于小波和分形理论的齿轮故障特征提取及噪声的和谐化研究.机械工程,重庆大学,2006,4(1):1-145.
48王庆华,张兴彪,张洪朋,孙玉清.分形理论在液压泵故障诊断中的应用.大连海事大学学报自然科学版,2004:40-43.
49蔡爱民,查良松.基于分形理论的安徽省旱、洪涝灾害时序特征分析.安徽农业大学学报,2005:546-550.
50孙金花,冯英浚,胡健.基于分形理论的股票时序数据离群模式挖掘研究.运筹与管理,2008:135-140.
51关腾,许娜.金融时间序列的多重分形分类.郑州大学学报(理学版),2008,4.
52 Raghav S,Mishra AK.Fractal feature based ECG arrh-ythmia classification.IEEE,Piscataway,NJ,USA,2008.5.
53钟维年,高清维,陈燕玲.基于小波和多重分形的金融时间序列聚类.系統工程,2009:58-61.
2 Mark AKHB.Last Data Mining In Time Series Databases,World Scientific,2004.
3韦东兴,陈晓云,徐荣聪.基于角点检测的图像形状特征提取方法.计算机工程,2010,4.
4加玉涛,罗志增.肌电信号特征提取方法综述.电子器件,2007:326-330.
5刘慧,谢洪波,卫星.基于模糊熵的脑电睡眠分期特征提取与分类.数据采集与处理,2010,25(4):484-489.
6 Greene BR,Boylan GB,Reilly RB.Connolly Combination ofEEG and ECG for improved automatic neonatal seizuredetection,Clin Neurophysiol 2007,118:1348-1359.
7 Gao GQ.Computerised detection and classification of fivecardiac conditions,AUT University.
8商卫波.心电信号自动分析与诊断处理方法研究.西安:西北工业大学,2005.
9葛丁飞,邵宇权,蒋惠忠.基于双导联Ecg和多变量回归模型的远程心电诊断算法研究.航天医学与医学工程,2004:355-359.
10 G.D.-F.H.B.-P.X.Xin-Jian.Study of feature extraction basedon autoregressive modeling in ECG automatic diagnosis.Acta Automatica Sinica,2007,33(5).
11葛丁飞,夏顺仁.Ar模型在远程心电诊断中的应用.中国生物医学工程学报,2004:222-229.
12钱争鸣.Arch族计量模型在金融市场研究中的应用.厦门大学学报(哲学社会科学版),2000:126-129.
13余素红,张世英.Sv与Garch模型对金融时间序列刻画能力的比较研究.系统工程,2002:28-33.
14 Kalantzis T,Papanastassiou D.Classification of GARCHtime series:an empirical investigation.Applied FinancialEconomics,2008,18:759-764.
15 Mrchen F.Time series feature extraction for data miningusing DWT and DFT,2003.
16 Valenza G,Lanata A,Ferro M,Scilingo EP.Real-timediscrimination of multiple cardiac arrhythmias for wearablesystems based on neural networks.Computers in Cardiology,2008.1053-1056.
17郭躬德,王晖.DavidBell时间序列数据分析与预处理.小型微型计算机系统,2003:2228-2232.
18江近仁,孙景江.以反应谱和功率谱密度函数表征的强震运动的统计特性.中国地震,1994:327-340.
19 Amini AM.Qualitative features extraction from sensor datausing Short-Time Fourier Transform,Visual Information Pr-ocessing XIV 5817(2005)289-295.
20 Holmberg M,Gelbart D,Hemmert W.Automatic speechrecognition with an adaptation model motivated by auditoryprocessing,Audio,Speech,and Language Processing.IEEETrans.on,2006,14:43-49.
21 Xu H,Tan ZH,Dalsgaard P,Lindberg B.Exploitation ofspectral variance to improve robustness in speech recogni-tion.Electronics Letters,2006,42:312-314.
22 F.H.a.H.Ney Noise Level Normalization and ReferenceAdaptation for Robust Speech Recognition.InternationalWorkshop on Automatic Speech Recognition:Challengesfor the New Millennium.2000.
23 Hilger F,Ney H.Quantile based histogram equalization fornoise robust large vocabulary speech recognition,Audio,Speech,and Language Processing.IEEE Trans.on,2006,14:845-854.
24高瑞华.多种预处理方法在语音检测中应用效果的比较研究.控制理论与控制工程[硕士学位论文].浙江工业大学,2004.64.
25 J.R.G.Townshend,C.O.Justice and V.Kalb Characterizationand Classification of South American Land Cover TypesUsing Satellite Data.International Journal of RemoteSensing,1987,8:1189-1207.
26邢玉娟,谭萍,李明.一种新的说话人识别序列特征提取方法.兰州理工大学学报,2009,35(4):98-102.
27 Scholkopf SAB,Muller KR.Nonlinear component analysis asa kernel eigenvalue problem.Neural Computaion.
28杨绍华.一种基于核主成分分析的人脸识别方法.河北科技师范学院学报,2008:45-48,62.
29周小程,马向玲,范洪达,庞文强.基于核函数分类的多维时序特征选择方法应用.电光与控制,2010,17(7):74-77.
30侯振雨,蔡文生,邵学广.主成分分析-支持向量回归建模方法及应用研究.分析化学,2006:617-620.
31楼蒋.Method of Feature Extraction Suitable for Hyperdi-mensional Time Series Data.浙江科技学院学报,2007,19(2).
32胡利平,刘宏伟,吴顺君.基于两级2dpca的Sar目标特征提取与识别.电子与信息学报,2008:1722-1726.
33李志农,曾明如,韩捷,何永勇.主分量分析和因子隐Markov模型在机械故障诊断中的应用.机械强度,2007:25-29.
34朱帮助.基于特征提取的选择性神经网络集成方法.计算机科学,2008:132-133,172.
35刘爱霞,刘正军,王静.基于Pca变换和神经元网络分类方法的中国森林制图研究.长江流域资源与环境,2006:19-24.
36王海燕,刘鲁,刘玲.基于Gra和pca的Bp神经网络应用研究.管理评论,2007:50-54.
37张静远,张冰.基于小波变换的特征提取方法分析.信号处理,2000:156-162,155.
38马陈,马国亮.心电图的小波变换特征提取.中国科技信息,2006(1A).
39 Andreao RV,Dorizzi B,Boudy J.ECG signal analysisthrough hidden Markov models.IEEE Trans.on BiomedicalEngineering,2006,53:1541-1549.
40胡博,卫宏儒,廖福成.小波变换在股票分析中的应用.系统管理学报,2007:365-369.
41金解放,赵奎,王晓军,赵康.岩石声发射信号处理小波基选择的研究.矿业研究与开发,2007:12-15.
42朱江梅,吴兴方.小波包变换压缩与地震数据处理.中国海上油气(工程),2006:169-173.
43田喜英.主元分析和非线性方法在表面肌电信号中的应用研究[硕士学位论文].上海交通大学.2009.
44 Gupta AMR,Singh K.A Time-Series-Based FeatureExtraction Approach for Prediction of Protein StructuralClass,EURASIP.Journal on Bioinformatics and SystemsBiology,2008.
45 Palancz B,Benyo B.Classification of time series viawavelet subband analysis using support vector machineclassifier.Periodica Polytechnica Electrical Engineering 50,2006:129-140.
46郑继明,邢峰,吴渝,李婧.基于小波变换和支持向量机的音频分类.重庆邮电大学学报(自然科学版),2008:212-216.
47王春.基于小波和分形理论的齿轮故障特征提取及噪声的和谐化研究.机械工程,重庆大学,2006,4(1):1-145.
48王庆华,张兴彪,张洪朋,孙玉清.分形理论在液压泵故障诊断中的应用.大连海事大学学报自然科学版,2004:40-43.
49蔡爱民,查良松.基于分形理论的安徽省旱、洪涝灾害时序特征分析.安徽农业大学学报,2005:546-550.
50孙金花,冯英浚,胡健.基于分形理论的股票时序数据离群模式挖掘研究.运筹与管理,2008:135-140.
51关腾,许娜.金融时间序列的多重分形分类.郑州大学学报(理学版),2008,4.
52 Raghav S,Mishra AK.Fractal feature based ECG arrh-ythmia classification.IEEE,Piscataway,NJ,USA,2008.5.
53钟维年,高清维,陈燕玲.基于小波和多重分形的金融时间序列聚类.系統工程,2009:58-61.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心