时间序列模式匹配技术研究
|
摘要
时间序列模式匹配是时间序列挖掘研究中一个基础和重要的问题，它是聚类、分类、规则发现、异常检测和知识发现等其它挖掘任务的基本技术手段和处理方法。该课题的研究目前还处在积极发展的阶段，现有的研究成果已经显示出模式匹配具有广阔的应用前景。
以目前时间序列的主体框架GEMINI框架为基础，开展时间序列模式匹配技术中的分类、子序列的定位、序列间距离的度量、距离度量的优化和适合于动态时间序列的匹配方法等的研究具有重要意义。
按照匹配的形式可以将模式匹配问题分为子序列匹配和全序列匹配两类。子序列匹配是从长度更长的主序列中查找与给定的模式序列相同或相似子序列的过程，最重要的操作是子序列定位，解决该问题的关键是如何减少匹配次数。全序列匹配是从时间序列数据库中查找与给定的模式序列相同或相似序列的过程。全序列匹配一般转换为距离度量的问题，如何提高序列之间距离度量的准确度和效率是该问题的研究热点。
通过对现有的经典子序列定位方法的深入分析,提出了基于重复度的贪心定位算法和基于相异度的贪心定位算法。贪心定位算法运用最优化原理计算模式序列的特征值，模式匹配按照特征值的大小从大到小进行，具有匹配效率高、适应性好、易于扩展和易于优化的优点。
针对全序列匹配中采用的距离度量方法存在度量准确度低、效率低的问题，提出了基于形态拟合的距离度量算法。该算法将欧式距离和动态弯曲距离两种度量方法进行优势组合，尽力消除影响欧式距离和动态弯曲距离度量的不利因素，具有度量准确度高和效率高的优点。
现有精确的动态弯曲距离存在计算性能瓶颈，在深入分析动态弯曲距离计算原理的基础上，提出了距离上界函数的概念，并给出了动态弯曲距离的距离上界函数的设计方法。如果函数的计算路径是一条动态弯曲路径，那么它就是一个动态弯曲距离的距离上界函数。为提高精确的动态弯曲距离的计算效率，提出了应用提前终止技术加速计算过程的新算法，该算法以动态弯曲距离的距离上界函数作为提前终止判定值，在计算累积距离矩阵的过程中，当累积距离大于动态弯曲距离的距离上界函数值时，对该计算路径进行终止标识，重新选取其它的计算路径进行计算。该算法确保了动态弯曲距离的度量准确度，同时有效地提高了计算效率。
在深入分析动态序列、动态匹配和匹配算法特点的基础上，提出了模式序列的模式阈值的概念，设计和实现了适合于动态时间序列的基于模式阈值的定位算法，并对该算法进行了优化。
以上述研究成果为基础，设计和实现了一个同时支持子序列匹配、全序列匹配和动态匹配的时间序列模式匹配原型系统TSPM(Time Series Pattern Matching),为评估各类匹配方法的实际效果提供了实验研究平台。
Time series pattern matching is an important and fundamental problem in time series datamining research. It is the basic technique and method used in clustering, classification, rulediscovery, abnormal detection, knowledge discovery and other data mining research tasks.Time series pattern matching is still an active research topic with existing research resultsshowing potential for widespread applications.
The GEMINI framework proposed by Faloutsos et al has become a mainstream frameworkin this research area. This dissertation models time series with GEMINI framework andrelated research is carried out based on this framework. To carry out the classification,sub-series localization, distance measure, the distance measure optimization and dynamicmatching method is important to study.
In this dissertation, we categorize time series pattern matching into sub-series matching andwhole-series matching. Sub-series matching is a process to search for a sub-series identical orsimilar to a given series from a longer main series. The key to the solution is to reduce thenumber of matching comparisons; the most important operation in sub-series matching is tolocalizing the sub-series. Whole-series matching is a process to search for a series identical orsimilar to a given series in a database of time series. Whole-series matching is usuallyconverted to distance measure problem and its key is to enhance the accuracy and efficiencyof estimating distance measure between time series.
Localizing sub-series is the most important operation in sub-series matching which haswide spread application. We propose a new greedy algorithm for sub-series localization basedon classic greedy algorithm. Our greedy algorithm calculates the eigenvalue of a time seriesusing the principle of optimality. Pattern matching carries on in the decreasing order ofeigenvalues. The algorithm is efficient and easy to adapt, extend and optimize.
Whole series matching problem is usually solved by converting to distance measureproblem. Existing distance measures suffer from either low accuracy or low efficiency. Wepropose a form-fitting distance measure which combines the advantages of both Euclideandistance and dynamic warping distance. It has the advantages of high accuracy and highefficiency.
Existing exact dynamic warping distance has a performance bottleneck. We propose theconcept and definition of distance upper bound function, analyze its principle of application,and provide methods to design upper bound function for dynamic warping distance. If the calculation path is a dynamic warping path, then it is an upper bound function of dynamicwarping distance. To enhance the efficiency of exact dynamic warping distance computation,we propose a new acceleration method based on early termination. The new method usesupper bound function of the dynamic warping distance as decision variable for earlytermination. In the process of cumulative distance matrix, a calculation path is flagged to beterminated when the cumulative distance becomes greater than the distance upper boundfunction, calculation will carry on along an alternative path. The proposed method guaranteesthe accuracy of dynamic warping distance while simultaneously enhance the calculationefficiency.
Dynamic time series are a series of observation values arrived in order. Pattern matchingfor dynamic time series is more difficult than static time series. Sub-series matching is themain operation in dynamic matching of which localizing sub-series is the most importantoperation. This dissertation proposes the concept of pattern threshold for pattern series. Wedesign and implement localization method based on pattern threshold for dynamic time seriesmatching.
Finally, we design and implement a prototype system for time series pattern matchingwhich supports sub-series matching, whole-series matching and dynamic time seriesmatching. This prototype system provides a research platform for evaluating variousmatching methods.
以目前时间序列的主体框架GEMINI框架为基础，开展时间序列模式匹配技术中的分类、子序列的定位、序列间距离的度量、距离度量的优化和适合于动态时间序列的匹配方法等的研究具有重要意义。
按照匹配的形式可以将模式匹配问题分为子序列匹配和全序列匹配两类。子序列匹配是从长度更长的主序列中查找与给定的模式序列相同或相似子序列的过程，最重要的操作是子序列定位，解决该问题的关键是如何减少匹配次数。全序列匹配是从时间序列数据库中查找与给定的模式序列相同或相似序列的过程。全序列匹配一般转换为距离度量的问题，如何提高序列之间距离度量的准确度和效率是该问题的研究热点。
通过对现有的经典子序列定位方法的深入分析,提出了基于重复度的贪心定位算法和基于相异度的贪心定位算法。贪心定位算法运用最优化原理计算模式序列的特征值，模式匹配按照特征值的大小从大到小进行，具有匹配效率高、适应性好、易于扩展和易于优化的优点。
针对全序列匹配中采用的距离度量方法存在度量准确度低、效率低的问题，提出了基于形态拟合的距离度量算法。该算法将欧式距离和动态弯曲距离两种度量方法进行优势组合，尽力消除影响欧式距离和动态弯曲距离度量的不利因素，具有度量准确度高和效率高的优点。
现有精确的动态弯曲距离存在计算性能瓶颈，在深入分析动态弯曲距离计算原理的基础上，提出了距离上界函数的概念，并给出了动态弯曲距离的距离上界函数的设计方法。如果函数的计算路径是一条动态弯曲路径，那么它就是一个动态弯曲距离的距离上界函数。为提高精确的动态弯曲距离的计算效率，提出了应用提前终止技术加速计算过程的新算法，该算法以动态弯曲距离的距离上界函数作为提前终止判定值，在计算累积距离矩阵的过程中，当累积距离大于动态弯曲距离的距离上界函数值时，对该计算路径进行终止标识，重新选取其它的计算路径进行计算。该算法确保了动态弯曲距离的度量准确度，同时有效地提高了计算效率。
在深入分析动态序列、动态匹配和匹配算法特点的基础上，提出了模式序列的模式阈值的概念，设计和实现了适合于动态时间序列的基于模式阈值的定位算法，并对该算法进行了优化。
以上述研究成果为基础，设计和实现了一个同时支持子序列匹配、全序列匹配和动态匹配的时间序列模式匹配原型系统TSPM(Time Series Pattern Matching),为评估各类匹配方法的实际效果提供了实验研究平台。
Time series pattern matching is an important and fundamental problem in time series datamining research. It is the basic technique and method used in clustering, classification, rulediscovery, abnormal detection, knowledge discovery and other data mining research tasks.Time series pattern matching is still an active research topic with existing research resultsshowing potential for widespread applications.
The GEMINI framework proposed by Faloutsos et al has become a mainstream frameworkin this research area. This dissertation models time series with GEMINI framework andrelated research is carried out based on this framework. To carry out the classification,sub-series localization, distance measure, the distance measure optimization and dynamicmatching method is important to study.
In this dissertation, we categorize time series pattern matching into sub-series matching andwhole-series matching. Sub-series matching is a process to search for a sub-series identical orsimilar to a given series from a longer main series. The key to the solution is to reduce thenumber of matching comparisons; the most important operation in sub-series matching is tolocalizing the sub-series. Whole-series matching is a process to search for a series identical orsimilar to a given series in a database of time series. Whole-series matching is usuallyconverted to distance measure problem and its key is to enhance the accuracy and efficiencyof estimating distance measure between time series.
Localizing sub-series is the most important operation in sub-series matching which haswide spread application. We propose a new greedy algorithm for sub-series localization basedon classic greedy algorithm. Our greedy algorithm calculates the eigenvalue of a time seriesusing the principle of optimality. Pattern matching carries on in the decreasing order ofeigenvalues. The algorithm is efficient and easy to adapt, extend and optimize.
Whole series matching problem is usually solved by converting to distance measureproblem. Existing distance measures suffer from either low accuracy or low efficiency. Wepropose a form-fitting distance measure which combines the advantages of both Euclideandistance and dynamic warping distance. It has the advantages of high accuracy and highefficiency.
Existing exact dynamic warping distance has a performance bottleneck. We propose theconcept and definition of distance upper bound function, analyze its principle of application,and provide methods to design upper bound function for dynamic warping distance. If the calculation path is a dynamic warping path, then it is an upper bound function of dynamicwarping distance. To enhance the efficiency of exact dynamic warping distance computation,we propose a new acceleration method based on early termination. The new method usesupper bound function of the dynamic warping distance as decision variable for earlytermination. In the process of cumulative distance matrix, a calculation path is flagged to beterminated when the cumulative distance becomes greater than the distance upper boundfunction, calculation will carry on along an alternative path. The proposed method guaranteesthe accuracy of dynamic warping distance while simultaneously enhance the calculationefficiency.
Dynamic time series are a series of observation values arrived in order. Pattern matchingfor dynamic time series is more difficult than static time series. Sub-series matching is themain operation in dynamic matching of which localizing sub-series is the most importantoperation. This dissertation proposes the concept of pattern threshold for pattern series. Wedesign and implement localization method based on pattern threshold for dynamic time seriesmatching.
Finally, we design and implement a prototype system for time series pattern matchingwhich supports sub-series matching, whole-series matching and dynamic time seriesmatching. This prototype system provides a research platform for evaluating variousmatching methods.
引文
[1] Douglas C Montgomery, Cheryl L Jennings, Murat Kulahci. Introduction to timeseries analysis and forecasting. Wiley Series in Probability and Statistics, John Wiley&Sons.2011.3~18
[2] Sakurai Y, Faloutsos C, Yamamuro M. Stream Monitoring under the Time WarpingDistance. in: Su Stanley Y W (Ed.). Proceedings of International Conference on DataEngineering (ICDE), Istanbul, Turkey,2007. IEEE Computer Society,2007.1046~1055
[3] Huijse P, Estevez PA, Zegers P, Principe JC. Period Estimation in Astronomical TimeSeries Using Slotted Correntropy. Signal Processing Letters, IEEE.2011,18(6):371~374
[4] JIA Ming-ming, LIU Dian-wei, SONG Kai-shan, WANG Zong-ming, JIANGGuang-jia, DU Jia, ZENG Li-hong. Classification and Verification of Land Use/Coverin Australia Using MODIS Time-series Data. Remote Sensing Technology andApplication2010,25(3):379~386
[5] Zhu Y, Shasha D. Warping indexes with envelope transforms for query by humming.in: Halevy A Y, Ives Z G, Doan A (Eds.). Proceedings of ACM SIGMODInternational Conference on Management of Data, San Diego, California, USA,2003.ACM Press,2003.181~192
[6]吴邵春,吴耿锋,王炜等.寻找地震相关地区的时间序列相似性匹配算法.软件学报,2006.27(3):185~192
[7] Chi-Jie Lu, Tian-Shyug Lee, Chih-Chou Chiu. Financial time series forecasting usingindependent component analysis and support vector regression. Decision SupportSystems,2009,47(2):115~125
[8] Keogh E, Kasetty S. On the need for time series data mining benchmarks: a surveyand empirical demonstration. Data Mining and Knowledge Discovery,2003,7(4):349~371
[9] Mehmed Kantardzic. Data Mining: Concepts, Models, Methods, and Algorithms. JohnWiley&Sons.2011.10~21
[10] Krzysztof J, Cios W P, Roman W, Swiniarski, Lukasz A K. Data Mining: AKnowledge Discovery Approach. Springer Publishing Company, Incorporated.2010.27~34
[11] Keogh E, Ratanamahatana C. Exact indexing of dynamic time warping. in:Knowledge and Information Systems,2005,7(3):358~386
[12] Keogh E. Similarity Search in Massive Time Series Databases:[Ph.D. Thesis].University of California, Irvine, USA.2002
[13] Yang Q, Wang X.10Challenging Problems in data mining research. InternationalJournal of Information Technology and Decision Making,2006,5(4):597~604
[14] Arne Koopman, Arno Knobbe, Marvin Meeng. Pattern selection problems inmultivariate time-series using equation discovery. Proceeding UP10Proceedings ofthe ACM SIGKDD Workshop on Useful Patterns. ACM New York, NY, USA.2010.74~81
[15] Agrawal R, Faloutsos C, Swami A. Efficient similarity search in sequence in sequencedatabases. in: Lomet D (Ed.). Proceedings of4th International Conference onFoundations of Data Organization and Algorithms, Chicago, Illinois, USA,1993.Springer Verlag,1994.69~84
[16] Molnar S, Terdik G. A General Fractal Model of Internet Traffic, in: Michaletzky CGy (Ed.). Proceedings of the26th Annual IEEE Conference on Local ComputerNetworks (LCN). Tampa, Florida, USA,2001. IEEE Computer Society,2001.492~499
[17]何书元.应用时间序列分析.北京:北京大学出版社,2004.1~15
[18] Zhu Y. High Performance Data Mining in Time Series: Techniques and Case Studies:
[Ph.D. Thesis], New York University, USA,2004.31~33
[19] Montani S, Bottrighi A, Leonardi G, Portinale L. A CBR-BASED, CLOSED-LOOPARCHITECTURE FOR TEMPORAL ABSTRACTIONS CONFIGURATION.Computational Intelligence.2009,25:235~249
[20] Chan K, Fu A W. Efficient time series matching by wavelets. in: Richard S T (Ed.).Proceedings of15th IEEE International Conference on Data Engineering (ICDE).Sydney, Australia,1999. IEEE Computer Society,1999.126~133
[21] Yu W, Chen G, Cao J, et al. Parameter identification of dynamic systems from timeseries. Physical Review E,2007,75(6):221~242
[22]李爱国,覃征.大规模时间序列数据库降维及相似搜索.计算机学报,2005,1(9):1447~1475
[23]黄超,朱扬勇.基于回归系数的时间序列维约简与相似性查找.模式识别与人工智能,2006,19(1):52~57
[24]肖辉.时间序列的相似性查询与异常检测:[博士学位论文].上海:复旦大学图书馆,2005.31~33
[25]王达,荣冈.时间序列的模式距离.浙江大学学报(工学版),2004,39(7):795~798
[26]李斌,谭立湘,章劲松等.面向数据挖掘的时间序列符号化方法研究.电路与系统学报,2000,5(2):9~14
[27]李爱国,覃征,贺升平.时间序列数据的相似模式抽取.西安交通大学学报,2002,36(2):1275~1278
[28] John F R, Myra S. A survey of temporal knowledge discovery paradigms and methods.IEEE Transactions on Knowledge and Data Engineering,2002,14(4):750~767
[29]潘定,沈钧毅.时态数据挖掘的相似性发现技术.软件学报,2006,18(2):246~258
[30] Faloutsos C, Ranganathan M, Manolopoulos Y. Fast subsequence matching intime-series databases. in: Snodgrass R T, Winslett M (Eds.). Proceedings of ACMSIGMOD Conference, Minneapolis, USA,1994. ACM Press,1994.419~429
[31] Shatkay H, Zdonik H S. Approximate queries and representations for large datasequences. in: Stanley Y, Su W (Eds.). Proceedings of12th International Conferenceon Data Engineering (ICDE), New Orleans, Louisiana, USA,1996. IEEE ComputerSociety,1996.536~545
[32] Keogh E, Pazzani M. A simple dimensionality reduction technique for fast similaritysearch in large time series databases. in: Terano T, Liu H, Chen AL (Eds.).Proceedings of4th Pacific-Asia Conference on Knowledge Discovery and DataMining, Kyoto, Japan,2000. Springer-Verlag,2000.122~133
[33] Yi B, Faloutsos C. Fast time sequence indexing for arbitrary Lp norms. in: Abbadi AE,Brodie ML, Chakravarthy S (Eds.). Proceedings of the26th International Conferenceon Very Large Databases (VLDB), Cairo, Egypt,2000. Morgan Kaufmann Publishers,2000.385~394
[34] Keogh E, Chakrabarti K, Pazzani M. Locally adaptive dimensionality reduction forindexing large time series databases. in: Aref W G (Ed.). Proceedings of ACMSIGMOD Conference on Management of Data, Santa Barbara, USA,2001. ACMPress,2001.151~162
[35] Yi-Leh Wu, Divyakant Agrawal, Amr El Abbadi. A comparison of DFT and DWTbased similarity search in time-series databases. Proceedings of the ninth internationalconference on Information and knowledge management, November06-11,2000,McLean, Virginia, United States.2000.488~495
[36] Shahabi C, Tian X, Zhao W. TSA-tree: a wavelet-based approach to improve theefficiency of multilevel surprise and trend queries. in: Gunther O, Lenz H J (Eds.).Proceedings of12th International Conference of Scientific and Statistical DatabaseManagement, Berlin, Germany,2000. IEEE Computer Society,2000.55~68
[37] Chan F K P, Fu A W C, Yu C. Haar wavelets for efficient similarity search oftime-series: with and without time warping. IEEE Transactions on Knowledge andData Engineering.2003,15(3):686~705
[38] Kahveci T, Singh A. Variable length queries for time series data. in: Urban J,Dasgupta P (Eds.). Proceedings of the17th International Conference on DataEngineering (ICDE), Heidelberg, Germany,2001. IEEE Computer Society,2001.273~282
[39] Korn F, Jagadish H, Faloutsos C. Efficiently supporting ad hoc queries in largedatasets of time sequences. in: Joan P (Ed.). Proceedings of ACM SIGMODInternational Conference on Management of Data, Tuescon, AZ. USA,1997. MorganKaufmann Publisher,1997.289~300
[40] Pratt K B, Fink E. Search for Patterns in Compressed Time Series. in: InternationalJournal of Image and Graphics,2001,2(1):89~106
[41] Yu D, Qi Y, Xu Y, et al. Kernel-SOM Based Visualization Financial Time SeriesForecasting. in: Santos J M, Zapico Z (Eds.). Proceedings of the1st InternationalConference on Innovative Computing, Information and Control, Beijing, China,2006.IEEE Computer Society,2006.470~473
[42] Zaki, Mohammed J. SPADE: An Efficient Algorithm for Mining Frequent Sequences.Machine Learning, Springer Netherlands.2001,42(1):31~60
[43] Ge X, Smyth P. Deformable Markov Model Templates for Time Series PatternMatching. in: Simeon S J, Osmar Z R (Eds.). Proceedings of the6th ACM SIGKDDInternational Conference on Knowledge Discovery and Data Mining. Boston, USA,2000. ACM Press,2000.81~90
[44] Nakamura K, Matsumoto M. Incremental learning of context free grammars. in:Adriaans P, Fernau H, van Zaanen M (Eds.). Proceedings of the6th InternationalColloquium Grammatical Inference (ICGI), Amsterdam, Holland,2002.Springer-Verlag,2002.174~184
[45] Chiu B, Keogh E, Lonardi S. Probabilistic Discovery of Time Series Motifs. in: LiseG, Ted S E, Pedro D (Eds.). Proceedings of the9th ACM SIGKDD InternationalConference on Knowledge Discovery and Data Mining (SIGKDD), Washington DC,USA,2003. ACM Press,2003.493~498
[46] Keogh E, Lonardi S, Ratanamahatana A C. Towards Parameter-Free Data Mining. in:Evangelos S, Han J, Fayyad U M (Eds.). Proceedings of the10th ACM KnowledgeDiscovery in Database (KDD), Seattle, WA. USA.2004. ACM Press,2004.206~215
[47] Lin J, Keogh E, Londardi S. A Symbolic Representation of Time Series, withImplications for Streaming Algorithms. in: Mohammed J Z, Charu A C (Eds.).Proceedings of the8th ACM SIGMOD workshop on Research (DMKD), San Diego,CA, USA.2003. ACM Press,2003.55~68
[48] Perng C S, Wang H, Zhang S. Landmark: A new model for similarity-based patternquerying in time series database. in: Young D C (Ed.). Proceedings of the16thInternational Conference on Data Engineering (ICDE). San Diego, USA,2000. IEEEComputer Society Press,2000.33~42
[49] Olivera A L, Silva JP M. Efficient algorithms for the inference of minimum sizeDFAs. Machine Learning,2001,44(7):93~119
[50] Jun-Ichi Aoe. Computer algorithms: string pattern matching strategies. IEEEComputer Society Press,1994.36~37
[51] Knuth D E, Morris J H, Pratt V R. Fast pattern matching in strings. SIAM J Comput.1977.323~350
[52] Karp Richard M, Rabin Michael O. Efficient randomized pattern-matching algorithms.IBM Journal of Research and Development31(1987).249~260
[53] Jicheng Meng, Wenbin Zhang. Volume measure in2DPCA-based face recognition.Pattern Recognition Letters. Volume28, Issue10,15July2007.1203~1208
[54] Mattausch, Omori, Fukae Koide, Gyoten. Fully-parallel pattern-matching engine withdynamic adaptability to Hamming or Manhattan distance. VLSI Circuits Digest ofTechnical Papers.2002.252~255
[55] Wu L, Faloutsos C, Sycara K. FALCON: feedback adaptive loop for content-basedretrieval. in: Amr A E, Michael B L, Sharma C (Eds.). Proceedings of the26thInternational Conference on Very Large Databases (VLDB), Cairo, Egypt,2000.Morgan Kaufmann Press,2000.296~306
[56] Keogh E, Palpanas T, Zordan V B. Indexing Large Human-Motion Databases. in:Mairo A N, Tarmer M, Donald K, Miller R J (Eds.). Proceedings of30th Conferenceof Very Large Databases(VLDB), Toronto, Canada,2004. Morgan Kaufmann Press,2004.780~791
[57] Ristad E S, Yianilos P N. Learning String Edit Distance, in: Douglas F H (Ed.).Proceedings of14th International Conference on Machine Learning (ICML),Nashville, TN, USA,1997. Morgan Kaufmann Press,1997.287~295
[58] Klve, Te-Tsung Lin, Shi-Chun Tsai, Wen-Guey Tzeng. Permutation Arrays Under theChebyshev Distance. IEEE Transactions on Information Theory.2010.2611~2617
[59] Berndt D J, Clifford J. Finding Patterns in Time Series: A Dynamic ProgrammingApproach. in: Weld D, Clancey B (Eds.). Advances in Knowledge Discovery andData Mining, AAAI/MIT, Oregon, Portland,1996. The MIT Press,1996.229~248
[60] Chen L, Ng R. On the marriage of Lp-norm and edit distance. in: Nascimento MA,zsu MT, Kossmann D (Eds.). Proceedings of30th International Conference on VeryLarge Databases (VLDB), Toronto, Canada,2004. Morgan Kaufmann Publishers,2004.792~801
[61] Sandeep Tata, Jignesh M, Patel. Estimating the selectivity of tf-idf based cosinesimilarity predicates. SIGMOD Rec.36,2,2007.7~12
[62] Badrul Sarwar, George Karypis, Joseph Konstan, John Riedl. Item-based collaborativefiltering recommendation algorithms. In Proceedings of the10th internationalconference on World Wide Web (WWW '01). ACM, New York, NY, USA.2001.285~295
[63] Cohen Israel, Huang Yiteng, Chen Jingdong, Benesty Jacob. Primary Title: NoiseReduction in Speech Processing. Springer Berlin Heidelberg,2009.125~127
[64] Guha S, Rastogi R, Shim K. ROCK: a robust clustering algorithm for categoricalattributes. Data Engineering,1999. Proceedings.15th International Conference onData Engineering,1999.512~521
[65] Das G, Gunopulos D, Mannila H. Finding similar time series. in: Komorowski J,Zytkow J (Eds.). Proceeding of1st European Symposium on Principles of DataMining and Knowledge Discovery (PKDD), Bergen, Norway,1997. Springer-Verlag,1997.88~100
[66] Vlachos M, Kollios G, Gunopulos D. Discovering similar multidimensionaltrajectories. in: Agrawal R, Dittrich K, Ngu A H (Eds.). Proceedings of18thInternational Conference on Data Engineering (ICDE), San Jose, CA, USA,2002.IEEE Computer Society,2002.673~684
[67] Vlachos M, Hadjieleftheriou M, Gunopulos D. Indexing multi-dimensionaltime-series with support for multiple distance measures. in: Getoor L, Senator T E(Eds.). Proceedings of ACM SIGKDD International Conference on KnowledgeDiscovery and Data Mining. Washington DC, USA,2003. ACM Press,2003.216~225
[68] Gehrke J, Korn F, Srivastava D. On computing correlated aggregates over continualdata streams. Proceedings of ACM SIGMOD Conference on Management of Data,Santa Barbara, USA,2001. ACM Press,2001.13~24
[69] Arasu A, Bahu S, Widom J. The CQL continuous query language: Semanticfoundations and query execution. The VLDB Journal,2006,15(2):121~142
[70] Zhu Y, Shasha D. Efficient elastic burst detection in data streams. in: Halevy A Y,Ives Z G, Doan A (Eds.). Proceedings of ACM SIGMOD International Conference onKnowledge Discovery and Data Mining (SIGMOD), San Diego, CA, USA,2003.ACM Press,2003.336~345
[71] Keogh E, Chiu H, Pazzani M. Segmenting time series: a survey and novel approach.in: Last M, Kandel A, Bunks H (Eds.). Data Mining in Time Series Databases. WorldScientific Publishing Company, Singapore,2004.1~21
[72] Kleinberg J. Bursty and hierarchical structure in streams. in: Proceedings of the8thInternational Conference on Knowledge discovery and data mining (KDD), Edmonton,Alberta, Canada,2002. ACM Press,2002.131~150
[73] Keogh E, Chu S, Hart D. An Online Algorithm for Segmenting Time Series. in:Macintosh A, Moulton M, Preece A (Eds.). Proceedings of IEEE InternationalConference on Data Mining (ICDM), CA, USA,2001. IEEE Computer Society Press,2001.289~296
[74] Palpanas T, Vlachos M, Keogh E. Online Amnesic Approximation of Streaming TimeSeries. Proceedings of the20th International Conference on Data Engineering (ICDE),Boston, USA,2004. IEEE Computer Society,2004.338~349
[75] Lin X, Lu H, Xu J. Continuously maintaining quantile summaries of the most recent Nelements over a data stream. Proceedings of the20th International Conference on DataEngineering (ICDE), Boston, USA,2004. IEEE Computer Society,2004.362~373
[76] Kwon D, Lee S. Indexing the current positions of moving objects using the lazyupdate R-tree. Simeon S J, Chabane D (Eds.). Proceedings of the3rd InternationalConference on Mobile Data Management (MDM), Singapore,2002. IEEE ComputerSociety,2002.113~120
[77] Lee M, Hsu W, Ensen C S. Supporting frequent updates in R-trees: A bottom-upapproach. in: Johann F C, Peter L C, Serge A (Eds.). Proceedings of the29thInternational Conference on Very Large Data Bases (VLDB), Berlin, Germany,2003.Morgan Kaufmann,2003.608~619
[78] Jiawei Han, Micheline著.范明,孟小峰译.数据挖掘概念与技术.机械工业出版社,2007.323~345
[79] Rothe I, Susse H, Voss K. The method of normalization to determine invariants.Pattern Analysis and Machine Intelligence, IEEE Transactions,1996.366~376
[80] Hiromitsu Yamada, Kazuhiko Yamamoto, Taiichi Saito. A nonlinear normalizationmethod for handprinted kanji character recognition-line density equalization,1990.1023~1029
[81] Bolstad B M, Irizarry R A, strand M A, Speed T P. A comparison of normalizationmethods for high density oligonucleotide array data based onvariance and bias.Bioinformatics,2003,19(2):185~193
[82] Duttweiler D L. Proportionate normalized least-mean-squares adaptation in echocancelers. Speech and Audio Processing, IEEE Transactions,2000.508~518
[83] Cowlishaw M F. Decimal floating-point: algorism for computers. ComputerArithmetic,2003. in: Proceedings.16th IEEE Symposium,2003.104~111
[84]王省富.样条函数及其应用.西安:西北工业大学出版社,1989.178~181
[85]马少陆.二次平滑数字滤波.热能动力工程,1991,(1).105~108
[86]刘鹏,雷蕾,张雪凤.缺失数据处理方法的比较研究.计算机科学.2004(3):155~156
[87]史晓霞,李军治.时间序列消噪算法研究.计算机工程与应用,2007,43(25):51~53
[88]徐长发,李国宽.实用小波方法.武汉:华中科技大学出版社,2004.172~175
[89] Vidakovic B, Lozoya C B. On time-dependent wavlet denosing. IEEE Trans on SignalProcessing.1998,46(9):2549~2551
[90]程正兴,译.时间序列分析的小波方法.北京:机械工业出版社,2005.205~207
[91]吕志民,翟绪圣.基于奇异谱的降噪方法及其在故障诊断技术中的应用.机械工程学报,1999,(3):85~88
[92] Cole R, Hariharan R, Paterson M S, Zwick U. Tighter lower bounds on the exactcomplexity of string matching. SIAM J Comput,24(1995).30~45
[93] Cole R, Hariharan R. Tighter bounds on the exact complexity of string matching. in33rd Symp Found Comp Sci Pittsburgh, Pennsylvania,24-27Oct1992, IEEE.600~609
[94] Zubair M, Wahab F, Hussain I, Ikram, M. Text scanning approach for exact stringmatching.2010International Conference on Networking and Information Technology(ICNIT).2010.118~122
[95] Boyer R S, Moore J S. A fast string searching algorithm. Commun Assoc Comput.Mach,1977.762~772
[96] Hume A, Sunday D. Fast string searching,Software-Practice&Experience,21(1991).1221~1248
[97] Lecroq T. Experimental results on string matching algorithms. Software-Practice&Experience,25(1995).727~765.
[98] Smyth B. Computing Patterns in Strings. Addison Wesley Pearson.2003.201~203
[99]余祥宣,崔国华,邹海明.计算机算法基础(第二版).武汉:华中科技大学出版社.2000,105~106
[100] Keogh E, Lin J, Fu A. HOT SAX: Finding the Most Unusual Time SeriesSubsequence. in: Garofalakis M, Gehrke J, Rastogi R (Eds.). Proceedings ofInternational Conference of Data Mining (ICDM), Houston, Texas, USA,2005. IEEEComputer Society,2005.226~233
[101] Keogh E, Li W, Xi X. LB_Keogh Supports Exact Indexing of Shapes under RotationInvariance with Arbitrary Representations and Distance Measures. in: Dayal U,Whang K, Lomet D B (Eds.). ACM Very Large Data Bases (VLDB), Seoul, Korea,2006. Morgan Kaufmann Publishers,2006.882~893
[102] Li W, Keogh E, Van H H. Atomic Wedgie: Efficient Query Filtering for StreamingTime Series. in: Han J, Wah B W, Raghavan V (Eds.). Proceedings of the5th IEEEInternational Conference on Data Mining (ICDM), Houston, Texas,2005. IEEEComputer Society,2005.490~497
[103] Sakurai Y, Yoshikawa M, Faloutsos C. FTW: Fast Similarity Search under the TimeWrapping Distance. in: Sorvari S, Toldi O (Eds.). Proceedings of24th ACMSIGMOD-SIGACT-SIGART Symposium on Principles of Database Systems (PODS),Maryland, USA,2005. ACM Press,2005.326~337
[2] Sakurai Y, Faloutsos C, Yamamuro M. Stream Monitoring under the Time WarpingDistance. in: Su Stanley Y W (Ed.). Proceedings of International Conference on DataEngineering (ICDE), Istanbul, Turkey,2007. IEEE Computer Society,2007.1046~1055
[3] Huijse P, Estevez PA, Zegers P, Principe JC. Period Estimation in Astronomical TimeSeries Using Slotted Correntropy. Signal Processing Letters, IEEE.2011,18(6):371~374
[4] JIA Ming-ming, LIU Dian-wei, SONG Kai-shan, WANG Zong-ming, JIANGGuang-jia, DU Jia, ZENG Li-hong. Classification and Verification of Land Use/Coverin Australia Using MODIS Time-series Data. Remote Sensing Technology andApplication2010,25(3):379~386
[5] Zhu Y, Shasha D. Warping indexes with envelope transforms for query by humming.in: Halevy A Y, Ives Z G, Doan A (Eds.). Proceedings of ACM SIGMODInternational Conference on Management of Data, San Diego, California, USA,2003.ACM Press,2003.181~192
[6]吴邵春,吴耿锋,王炜等.寻找地震相关地区的时间序列相似性匹配算法.软件学报,2006.27(3):185~192
[7] Chi-Jie Lu, Tian-Shyug Lee, Chih-Chou Chiu. Financial time series forecasting usingindependent component analysis and support vector regression. Decision SupportSystems,2009,47(2):115~125
[8] Keogh E, Kasetty S. On the need for time series data mining benchmarks: a surveyand empirical demonstration. Data Mining and Knowledge Discovery,2003,7(4):349~371
[9] Mehmed Kantardzic. Data Mining: Concepts, Models, Methods, and Algorithms. JohnWiley&Sons.2011.10~21
[10] Krzysztof J, Cios W P, Roman W, Swiniarski, Lukasz A K. Data Mining: AKnowledge Discovery Approach. Springer Publishing Company, Incorporated.2010.27~34
[11] Keogh E, Ratanamahatana C. Exact indexing of dynamic time warping. in:Knowledge and Information Systems,2005,7(3):358~386
[12] Keogh E. Similarity Search in Massive Time Series Databases:[Ph.D. Thesis].University of California, Irvine, USA.2002
[13] Yang Q, Wang X.10Challenging Problems in data mining research. InternationalJournal of Information Technology and Decision Making,2006,5(4):597~604
[14] Arne Koopman, Arno Knobbe, Marvin Meeng. Pattern selection problems inmultivariate time-series using equation discovery. Proceeding UP10Proceedings ofthe ACM SIGKDD Workshop on Useful Patterns. ACM New York, NY, USA.2010.74~81
[15] Agrawal R, Faloutsos C, Swami A. Efficient similarity search in sequence in sequencedatabases. in: Lomet D (Ed.). Proceedings of4th International Conference onFoundations of Data Organization and Algorithms, Chicago, Illinois, USA,1993.Springer Verlag,1994.69~84
[16] Molnar S, Terdik G. A General Fractal Model of Internet Traffic, in: Michaletzky CGy (Ed.). Proceedings of the26th Annual IEEE Conference on Local ComputerNetworks (LCN). Tampa, Florida, USA,2001. IEEE Computer Society,2001.492~499
[17]何书元.应用时间序列分析.北京:北京大学出版社,2004.1~15
[18] Zhu Y. High Performance Data Mining in Time Series: Techniques and Case Studies:
[Ph.D. Thesis], New York University, USA,2004.31~33
[19] Montani S, Bottrighi A, Leonardi G, Portinale L. A CBR-BASED, CLOSED-LOOPARCHITECTURE FOR TEMPORAL ABSTRACTIONS CONFIGURATION.Computational Intelligence.2009,25:235~249
[20] Chan K, Fu A W. Efficient time series matching by wavelets. in: Richard S T (Ed.).Proceedings of15th IEEE International Conference on Data Engineering (ICDE).Sydney, Australia,1999. IEEE Computer Society,1999.126~133
[21] Yu W, Chen G, Cao J, et al. Parameter identification of dynamic systems from timeseries. Physical Review E,2007,75(6):221~242
[22]李爱国,覃征.大规模时间序列数据库降维及相似搜索.计算机学报,2005,1(9):1447~1475
[23]黄超,朱扬勇.基于回归系数的时间序列维约简与相似性查找.模式识别与人工智能,2006,19(1):52~57
[24]肖辉.时间序列的相似性查询与异常检测:[博士学位论文].上海:复旦大学图书馆,2005.31~33
[25]王达,荣冈.时间序列的模式距离.浙江大学学报(工学版),2004,39(7):795~798
[26]李斌,谭立湘,章劲松等.面向数据挖掘的时间序列符号化方法研究.电路与系统学报,2000,5(2):9~14
[27]李爱国,覃征,贺升平.时间序列数据的相似模式抽取.西安交通大学学报,2002,36(2):1275~1278
[28] John F R, Myra S. A survey of temporal knowledge discovery paradigms and methods.IEEE Transactions on Knowledge and Data Engineering,2002,14(4):750~767
[29]潘定,沈钧毅.时态数据挖掘的相似性发现技术.软件学报,2006,18(2):246~258
[30] Faloutsos C, Ranganathan M, Manolopoulos Y. Fast subsequence matching intime-series databases. in: Snodgrass R T, Winslett M (Eds.). Proceedings of ACMSIGMOD Conference, Minneapolis, USA,1994. ACM Press,1994.419~429
[31] Shatkay H, Zdonik H S. Approximate queries and representations for large datasequences. in: Stanley Y, Su W (Eds.). Proceedings of12th International Conferenceon Data Engineering (ICDE), New Orleans, Louisiana, USA,1996. IEEE ComputerSociety,1996.536~545
[32] Keogh E, Pazzani M. A simple dimensionality reduction technique for fast similaritysearch in large time series databases. in: Terano T, Liu H, Chen AL (Eds.).Proceedings of4th Pacific-Asia Conference on Knowledge Discovery and DataMining, Kyoto, Japan,2000. Springer-Verlag,2000.122~133
[33] Yi B, Faloutsos C. Fast time sequence indexing for arbitrary Lp norms. in: Abbadi AE,Brodie ML, Chakravarthy S (Eds.). Proceedings of the26th International Conferenceon Very Large Databases (VLDB), Cairo, Egypt,2000. Morgan Kaufmann Publishers,2000.385~394
[34] Keogh E, Chakrabarti K, Pazzani M. Locally adaptive dimensionality reduction forindexing large time series databases. in: Aref W G (Ed.). Proceedings of ACMSIGMOD Conference on Management of Data, Santa Barbara, USA,2001. ACMPress,2001.151~162
[35] Yi-Leh Wu, Divyakant Agrawal, Amr El Abbadi. A comparison of DFT and DWTbased similarity search in time-series databases. Proceedings of the ninth internationalconference on Information and knowledge management, November06-11,2000,McLean, Virginia, United States.2000.488~495
[36] Shahabi C, Tian X, Zhao W. TSA-tree: a wavelet-based approach to improve theefficiency of multilevel surprise and trend queries. in: Gunther O, Lenz H J (Eds.).Proceedings of12th International Conference of Scientific and Statistical DatabaseManagement, Berlin, Germany,2000. IEEE Computer Society,2000.55~68
[37] Chan F K P, Fu A W C, Yu C. Haar wavelets for efficient similarity search oftime-series: with and without time warping. IEEE Transactions on Knowledge andData Engineering.2003,15(3):686~705
[38] Kahveci T, Singh A. Variable length queries for time series data. in: Urban J,Dasgupta P (Eds.). Proceedings of the17th International Conference on DataEngineering (ICDE), Heidelberg, Germany,2001. IEEE Computer Society,2001.273~282
[39] Korn F, Jagadish H, Faloutsos C. Efficiently supporting ad hoc queries in largedatasets of time sequences. in: Joan P (Ed.). Proceedings of ACM SIGMODInternational Conference on Management of Data, Tuescon, AZ. USA,1997. MorganKaufmann Publisher,1997.289~300
[40] Pratt K B, Fink E. Search for Patterns in Compressed Time Series. in: InternationalJournal of Image and Graphics,2001,2(1):89~106
[41] Yu D, Qi Y, Xu Y, et al. Kernel-SOM Based Visualization Financial Time SeriesForecasting. in: Santos J M, Zapico Z (Eds.). Proceedings of the1st InternationalConference on Innovative Computing, Information and Control, Beijing, China,2006.IEEE Computer Society,2006.470~473
[42] Zaki, Mohammed J. SPADE: An Efficient Algorithm for Mining Frequent Sequences.Machine Learning, Springer Netherlands.2001,42(1):31~60
[43] Ge X, Smyth P. Deformable Markov Model Templates for Time Series PatternMatching. in: Simeon S J, Osmar Z R (Eds.). Proceedings of the6th ACM SIGKDDInternational Conference on Knowledge Discovery and Data Mining. Boston, USA,2000. ACM Press,2000.81~90
[44] Nakamura K, Matsumoto M. Incremental learning of context free grammars. in:Adriaans P, Fernau H, van Zaanen M (Eds.). Proceedings of the6th InternationalColloquium Grammatical Inference (ICGI), Amsterdam, Holland,2002.Springer-Verlag,2002.174~184
[45] Chiu B, Keogh E, Lonardi S. Probabilistic Discovery of Time Series Motifs. in: LiseG, Ted S E, Pedro D (Eds.). Proceedings of the9th ACM SIGKDD InternationalConference on Knowledge Discovery and Data Mining (SIGKDD), Washington DC,USA,2003. ACM Press,2003.493~498
[46] Keogh E, Lonardi S, Ratanamahatana A C. Towards Parameter-Free Data Mining. in:Evangelos S, Han J, Fayyad U M (Eds.). Proceedings of the10th ACM KnowledgeDiscovery in Database (KDD), Seattle, WA. USA.2004. ACM Press,2004.206~215
[47] Lin J, Keogh E, Londardi S. A Symbolic Representation of Time Series, withImplications for Streaming Algorithms. in: Mohammed J Z, Charu A C (Eds.).Proceedings of the8th ACM SIGMOD workshop on Research (DMKD), San Diego,CA, USA.2003. ACM Press,2003.55~68
[48] Perng C S, Wang H, Zhang S. Landmark: A new model for similarity-based patternquerying in time series database. in: Young D C (Ed.). Proceedings of the16thInternational Conference on Data Engineering (ICDE). San Diego, USA,2000. IEEEComputer Society Press,2000.33~42
[49] Olivera A L, Silva JP M. Efficient algorithms for the inference of minimum sizeDFAs. Machine Learning,2001,44(7):93~119
[50] Jun-Ichi Aoe. Computer algorithms: string pattern matching strategies. IEEEComputer Society Press,1994.36~37
[51] Knuth D E, Morris J H, Pratt V R. Fast pattern matching in strings. SIAM J Comput.1977.323~350
[52] Karp Richard M, Rabin Michael O. Efficient randomized pattern-matching algorithms.IBM Journal of Research and Development31(1987).249~260
[53] Jicheng Meng, Wenbin Zhang. Volume measure in2DPCA-based face recognition.Pattern Recognition Letters. Volume28, Issue10,15July2007.1203~1208
[54] Mattausch, Omori, Fukae Koide, Gyoten. Fully-parallel pattern-matching engine withdynamic adaptability to Hamming or Manhattan distance. VLSI Circuits Digest ofTechnical Papers.2002.252~255
[55] Wu L, Faloutsos C, Sycara K. FALCON: feedback adaptive loop for content-basedretrieval. in: Amr A E, Michael B L, Sharma C (Eds.). Proceedings of the26thInternational Conference on Very Large Databases (VLDB), Cairo, Egypt,2000.Morgan Kaufmann Press,2000.296~306
[56] Keogh E, Palpanas T, Zordan V B. Indexing Large Human-Motion Databases. in:Mairo A N, Tarmer M, Donald K, Miller R J (Eds.). Proceedings of30th Conferenceof Very Large Databases(VLDB), Toronto, Canada,2004. Morgan Kaufmann Press,2004.780~791
[57] Ristad E S, Yianilos P N. Learning String Edit Distance, in: Douglas F H (Ed.).Proceedings of14th International Conference on Machine Learning (ICML),Nashville, TN, USA,1997. Morgan Kaufmann Press,1997.287~295
[58] Klve, Te-Tsung Lin, Shi-Chun Tsai, Wen-Guey Tzeng. Permutation Arrays Under theChebyshev Distance. IEEE Transactions on Information Theory.2010.2611~2617
[59] Berndt D J, Clifford J. Finding Patterns in Time Series: A Dynamic ProgrammingApproach. in: Weld D, Clancey B (Eds.). Advances in Knowledge Discovery andData Mining, AAAI/MIT, Oregon, Portland,1996. The MIT Press,1996.229~248
[60] Chen L, Ng R. On the marriage of Lp-norm and edit distance. in: Nascimento MA,zsu MT, Kossmann D (Eds.). Proceedings of30th International Conference on VeryLarge Databases (VLDB), Toronto, Canada,2004. Morgan Kaufmann Publishers,2004.792~801
[61] Sandeep Tata, Jignesh M, Patel. Estimating the selectivity of tf-idf based cosinesimilarity predicates. SIGMOD Rec.36,2,2007.7~12
[62] Badrul Sarwar, George Karypis, Joseph Konstan, John Riedl. Item-based collaborativefiltering recommendation algorithms. In Proceedings of the10th internationalconference on World Wide Web (WWW '01). ACM, New York, NY, USA.2001.285~295
[63] Cohen Israel, Huang Yiteng, Chen Jingdong, Benesty Jacob. Primary Title: NoiseReduction in Speech Processing. Springer Berlin Heidelberg,2009.125~127
[64] Guha S, Rastogi R, Shim K. ROCK: a robust clustering algorithm for categoricalattributes. Data Engineering,1999. Proceedings.15th International Conference onData Engineering,1999.512~521
[65] Das G, Gunopulos D, Mannila H. Finding similar time series. in: Komorowski J,Zytkow J (Eds.). Proceeding of1st European Symposium on Principles of DataMining and Knowledge Discovery (PKDD), Bergen, Norway,1997. Springer-Verlag,1997.88~100
[66] Vlachos M, Kollios G, Gunopulos D. Discovering similar multidimensionaltrajectories. in: Agrawal R, Dittrich K, Ngu A H (Eds.). Proceedings of18thInternational Conference on Data Engineering (ICDE), San Jose, CA, USA,2002.IEEE Computer Society,2002.673~684
[67] Vlachos M, Hadjieleftheriou M, Gunopulos D. Indexing multi-dimensionaltime-series with support for multiple distance measures. in: Getoor L, Senator T E(Eds.). Proceedings of ACM SIGKDD International Conference on KnowledgeDiscovery and Data Mining. Washington DC, USA,2003. ACM Press,2003.216~225
[68] Gehrke J, Korn F, Srivastava D. On computing correlated aggregates over continualdata streams. Proceedings of ACM SIGMOD Conference on Management of Data,Santa Barbara, USA,2001. ACM Press,2001.13~24
[69] Arasu A, Bahu S, Widom J. The CQL continuous query language: Semanticfoundations and query execution. The VLDB Journal,2006,15(2):121~142
[70] Zhu Y, Shasha D. Efficient elastic burst detection in data streams. in: Halevy A Y,Ives Z G, Doan A (Eds.). Proceedings of ACM SIGMOD International Conference onKnowledge Discovery and Data Mining (SIGMOD), San Diego, CA, USA,2003.ACM Press,2003.336~345
[71] Keogh E, Chiu H, Pazzani M. Segmenting time series: a survey and novel approach.in: Last M, Kandel A, Bunks H (Eds.). Data Mining in Time Series Databases. WorldScientific Publishing Company, Singapore,2004.1~21
[72] Kleinberg J. Bursty and hierarchical structure in streams. in: Proceedings of the8thInternational Conference on Knowledge discovery and data mining (KDD), Edmonton,Alberta, Canada,2002. ACM Press,2002.131~150
[73] Keogh E, Chu S, Hart D. An Online Algorithm for Segmenting Time Series. in:Macintosh A, Moulton M, Preece A (Eds.). Proceedings of IEEE InternationalConference on Data Mining (ICDM), CA, USA,2001. IEEE Computer Society Press,2001.289~296
[74] Palpanas T, Vlachos M, Keogh E. Online Amnesic Approximation of Streaming TimeSeries. Proceedings of the20th International Conference on Data Engineering (ICDE),Boston, USA,2004. IEEE Computer Society,2004.338~349
[75] Lin X, Lu H, Xu J. Continuously maintaining quantile summaries of the most recent Nelements over a data stream. Proceedings of the20th International Conference on DataEngineering (ICDE), Boston, USA,2004. IEEE Computer Society,2004.362~373
[76] Kwon D, Lee S. Indexing the current positions of moving objects using the lazyupdate R-tree. Simeon S J, Chabane D (Eds.). Proceedings of the3rd InternationalConference on Mobile Data Management (MDM), Singapore,2002. IEEE ComputerSociety,2002.113~120
[77] Lee M, Hsu W, Ensen C S. Supporting frequent updates in R-trees: A bottom-upapproach. in: Johann F C, Peter L C, Serge A (Eds.). Proceedings of the29thInternational Conference on Very Large Data Bases (VLDB), Berlin, Germany,2003.Morgan Kaufmann,2003.608~619
[78] Jiawei Han, Micheline著.范明,孟小峰译.数据挖掘概念与技术.机械工业出版社,2007.323~345
[79] Rothe I, Susse H, Voss K. The method of normalization to determine invariants.Pattern Analysis and Machine Intelligence, IEEE Transactions,1996.366~376
[80] Hiromitsu Yamada, Kazuhiko Yamamoto, Taiichi Saito. A nonlinear normalizationmethod for handprinted kanji character recognition-line density equalization,1990.1023~1029
[81] Bolstad B M, Irizarry R A, strand M A, Speed T P. A comparison of normalizationmethods for high density oligonucleotide array data based onvariance and bias.Bioinformatics,2003,19(2):185~193
[82] Duttweiler D L. Proportionate normalized least-mean-squares adaptation in echocancelers. Speech and Audio Processing, IEEE Transactions,2000.508~518
[83] Cowlishaw M F. Decimal floating-point: algorism for computers. ComputerArithmetic,2003. in: Proceedings.16th IEEE Symposium,2003.104~111
[84]王省富.样条函数及其应用.西安:西北工业大学出版社,1989.178~181
[85]马少陆.二次平滑数字滤波.热能动力工程,1991,(1).105~108
[86]刘鹏,雷蕾,张雪凤.缺失数据处理方法的比较研究.计算机科学.2004(3):155~156
[87]史晓霞,李军治.时间序列消噪算法研究.计算机工程与应用,2007,43(25):51~53
[88]徐长发,李国宽.实用小波方法.武汉:华中科技大学出版社,2004.172~175
[89] Vidakovic B, Lozoya C B. On time-dependent wavlet denosing. IEEE Trans on SignalProcessing.1998,46(9):2549~2551
[90]程正兴,译.时间序列分析的小波方法.北京:机械工业出版社,2005.205~207
[91]吕志民,翟绪圣.基于奇异谱的降噪方法及其在故障诊断技术中的应用.机械工程学报,1999,(3):85~88
[92] Cole R, Hariharan R, Paterson M S, Zwick U. Tighter lower bounds on the exactcomplexity of string matching. SIAM J Comput,24(1995).30~45
[93] Cole R, Hariharan R. Tighter bounds on the exact complexity of string matching. in33rd Symp Found Comp Sci Pittsburgh, Pennsylvania,24-27Oct1992, IEEE.600~609
[94] Zubair M, Wahab F, Hussain I, Ikram, M. Text scanning approach for exact stringmatching.2010International Conference on Networking and Information Technology(ICNIT).2010.118~122
[95] Boyer R S, Moore J S. A fast string searching algorithm. Commun Assoc Comput.Mach,1977.762~772
[96] Hume A, Sunday D. Fast string searching,Software-Practice&Experience,21(1991).1221~1248
[97] Lecroq T. Experimental results on string matching algorithms. Software-Practice&Experience,25(1995).727~765.
[98] Smyth B. Computing Patterns in Strings. Addison Wesley Pearson.2003.201~203
[99]余祥宣,崔国华,邹海明.计算机算法基础(第二版).武汉:华中科技大学出版社.2000,105~106
[100] Keogh E, Lin J, Fu A. HOT SAX: Finding the Most Unusual Time SeriesSubsequence. in: Garofalakis M, Gehrke J, Rastogi R (Eds.). Proceedings ofInternational Conference of Data Mining (ICDM), Houston, Texas, USA,2005. IEEEComputer Society,2005.226~233
[101] Keogh E, Li W, Xi X. LB_Keogh Supports Exact Indexing of Shapes under RotationInvariance with Arbitrary Representations and Distance Measures. in: Dayal U,Whang K, Lomet D B (Eds.). ACM Very Large Data Bases (VLDB), Seoul, Korea,2006. Morgan Kaufmann Publishers,2006.882~893
[102] Li W, Keogh E, Van H H. Atomic Wedgie: Efficient Query Filtering for StreamingTime Series. in: Han J, Wah B W, Raghavan V (Eds.). Proceedings of the5th IEEEInternational Conference on Data Mining (ICDM), Houston, Texas,2005. IEEEComputer Society,2005.490~497
[103] Sakurai Y, Yoshikawa M, Faloutsos C. FTW: Fast Similarity Search under the TimeWrapping Distance. in: Sorvari S, Toldi O (Eds.). Proceedings of24th ACMSIGMOD-SIGACT-SIGART Symposium on Principles of Database Systems (PODS),Maryland, USA,2005. ACM Press,2005.326~337