E. J. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining (KDD'97), 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....and maxima. The search for a minimum cost state sequence in the automata of Section 2 and 4 can also be viewed as a search for approximate level sets in a time series, and hence related to the large body of work on piece wise function approximation in both statistics and data mining (see e.g. [26, 27, 30, 34, 36, 38, 43, 45]) In a discrete framework, work on mining episodes and sequential patterns (e.g. 1, 13, 28, 44] has developed algorithms to identify particular configurations of discrete events clustered in time, in some cases obeying partial precedence constraints on their order. Finally, there is an ....

E. Keogh, P. Smyth, "A probabilistic approach to fast pattern matching in time series databases," Proc. Intl. Conf. Knowledge Discovery and Data Mining, 1997.

....June 2002. An abbreviated version of this report will appear in Pvceedings of the 28th VLDB Confeence, Hong Kong, China, 2002 In mission operations for NASA s Space Shuttle, approximately 20,000 sensors are telemetered once per second to Mission Control at Johnson Space Center, Houston[16]. There are about 50,000 securities trading in the United States, and every second up to 100,000 quotes and trades (ticks) are generated. Unfortunately it is difficult to process such data in set oriented data management systems, although object relational time series extensions have begun to ....

E. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In the third conference on Knowledge Discovery in Databases and Data Mining, 1997.

....and maxima. The search for a minimum cost state sequence in the automata of Section 2 and 4 can also be viewed as a search for approximate level sets in a time series, and hence related to the large body of work on piece wise function approximation in both statistics and data mining (see e.g. [23, 24, 27, 31, 33, 35, 41]) In a discrete framework, work on mining episodes and sequential patterns (e.g. 1, 12, 25, 40] has developed algorithms to identify particular configurations of discrete events clustered in time, in some cases obeying partial precedence constraints on their order. Finally, there is an ....

E. Keogh, P. Smyth, "A probabilistic approach to fast pattern matching in time series databases," Proc. Intl. Conf. Knowledge Discovery and Data Mining, 1997.

....PC. The algorithm is embarrassingly parallelizable. 1 Introduction Many applications consist of multiple data streams. For example, In mission operations for NASA s Space Shuttle, approximately 20,000 sensors are telemetered once per second to Mission Control at Johnson Space Center, Houston[15]. There are about 50,000 securities trading in the United States, and every second up to 100,000 quotes and trades (ticks) are generated. Unfortunately it is dicult to process such data in set oriented data management systems, though object relational time series extensions have begun to ll ....

E. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In the third conference on Knowledge Discovery in Databases and Data Mining, 1997.

....and maxima. The search for a minimum cost state sequence in the automata of Section 2 and 4 can also be viewed as a search for approximate level sets in a time series, and hence related to the large body of work on piece wise function approximation in both statistics and data mining (see e.g. [24, 25, 28, 32, 34, 36, 41, 43]) In a discrete framework, work on mining episodes and sequential patterns (e.g. 1, 12, 26, 42] has developed algorithms to identify particular configurations of discrete events clustered in time, in some cases obeying partial precedence constraints on their order. Finally, there is an ....

E. Keogh, P. Smyth, "A probabilistic approach to fast pattern matching in time series databases," Proc. Intl. Conf. Knowledge Discovery and Data Mining, 1997.

.... handle multivariate data with a time component(while we have only visualised two of the fields over time, there are other time varying fields of interest too) An important area of current data mining research is the development of algorithms and visualisation techniques for time series analysis[5] and event sequence analysis[2] 8 4.2 Service Lags A service lag is the time interval between date of referral (DOR) and date of service (DOS ) for a pathology test. The summary of service lags in figure 3 reveals that most tests are ordered within 6 months of the referral date. This compares ....

E. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In Third Int. Conf. on Knowledge Discovery and Data Mining, pages 24--30. AAAI Press, Menlo Park,CA., 1997.

....also shown that this technique, like the PCA, can handle arbitrary L p norms. The empirical data suggests that the APCA outperforms both methods based on the discrete Fourier transform, and methods based on the discrete wavelet transform by one to two orders of magnitude. 3. 3 Landmark Methods In [20] Keogh and Smyth introduce a probabilistic method for sequence retrieval, where the features extracted are characteristic parts of the sequence, so called feature shapes. In [13] Keogh uses a similar landmark based technique. Both these methods also use dimensionality reduction technique of ....

....sequence element into one of a finite set of categories. Both methods achieve subquadratic running times but may not scale well for large time sequence databases. Keogh et al. have introduced a dimensionality reduction technique using piecewise linear approximation of the original sequence data [13, 16 18, 20]. This reduces the number of data points by a compression factor c, typically in the range from 10 to 600 for real data [13] outperforming methods based on the Discrete Fourier Transform by one to three orders of magnitude [18] This approximation is shown to be valid under several distance ....

Eamonn J. Keogh and Padhraic Smyth. A probabilistic approach to fast pattern matching in time series databases. page 24.

....the direct use of Fourier coe#cients [20] However, many sequences, in particular those containing transient behavior, are quire non stationary and may possess very weak spectral signatures even locally. Furthermore, from a knowledge discovery viewpoint, the spectral methods are somewhat indirect [37]. 3.2.5 Matching similar time series patterns Discovering common sequences and distinctive sequences has been given a great deal of attention ( 10, 51, 52, 19, 46, 88] Finding similar patterns in time series is used in several applications, including indexing like 11 patterns, finding ....

....46, 88] Finding similar patterns in time series is used in several applications, including indexing like 11 patterns, finding subsequences that are similar, clustering, and finding rules associating time series . Time series similarity has been approached in many di#erent ways. see e.g. [2, 9, 26, 20, 6, 17, 37, 10, 11, 13, 53]) Many approaches assume a template pattern (either global or local) and then try to find similar patterns in a reference sequence, see e.g. 2, 6, 26, 20, 74, 39] Most of the approaches decompose the time series into windows, and features are extracted from each window, and then e#cient ....

[Article contains additional citation context not shown here]

Keogh, E., Smyth, P., "A Probabilistic Approach to Fast Pattern Matching in Time Series Databases", Proceedings of the Third International Conference on Knowledge Discovery and Data Mining [KDD 97] , Newport Beach, California, 1997. 37

....have suggested abandoning the insistence on exact search in favor of a much faster search that returns approximately the same results. Typically this involves transforming the data with a lossy compression scheme, then doing a sequential search on the compressed data. Typical examples include [42, 27, 30, 46], who all utilize a piecewise linear approximation. Others have suggested transforming the data into a discrete alphabet and using string matching algorithms [2, 20, 34, 29, 21, 38] All these approaches suffer from some limitations. They are all evaluated on small datasets residing in main ....

....combining 7 datasets with widely varying properties of shape, structure, noise etc. The only preprocessing performed was to insure that each time series had a mean of zero and a standard deviation of one (otherwise many queries become pathologically easy) The 7 datasets are, Space Shuttle STS57 [27, 25], Arrhythmia [32] Random Walk [46, 34, 52, 24] InTERBALL Plasma processes (figure 4) 43] Astrophysical data (figure 1) 47] Pseudo Periodic Synthetic Time Series [4] Exchange rate (figure 4) 47] Once again, we generated data of 3 different dimensionalities: n=1024, n=512 and n=256 and in ....

Keogh, E., & Smyth, P. (1997). A probabilistic approach to fast pattern matching in time series databases. Proceedings of the 3 rd International Conference of Knowledge Discovery and Data Mining. pp 24-20.

....solutions. Several high level representations of time series have been proposed, including Fourier Transforms [1,13] Wavelets [4] Symbolic Mappings [2, 5, 24] and Piecewise Linear Representation (PLR) In this work, we confine our attention to PLR, perhaps the most frequently used representation [8, 10, 12, 14, 15, 16, 17, 18, 20, 21, 22, 25, 27, 28, 30, 31]. Intuitively Piecewise Linear Representation refers to the approximation of a time series T, of length n, with K straight lines. Figure 1 contains two examples. Because K is typically much smaller that n, this representation makes the storage, transmission and computation of the data more ....

.... algorithm Two and three dimensional analogues of this algorithm are common in the field of computer graphics where they are called decimation methods [9] In data mining, the algorithm has been used extensively by two of the current authors to support a variety of time series data mining tasks [14, 15, 16]. In medicine, the algorithm was used by Hunter and McIntosh to provide the high level representation for their medical pattern matching system [10] 2.4 Feature Comparison of the Major Algorithms We have deliberately deferred the discussion of the running times of the algorithms until now, when ....

[Article contains additional citation context not shown here]

Keogh, E., & Smyth, P. (1997). A probabilistic approach to fast pattern matching in time series databases. Proceedings of the 3 International Conference of Knowledge Discovery and Data Mining. pp 24-20.

.... depends on the structure of the data itself and the task at hand (i.e. clustering classification retrieval etc) For most applications the best approach may be to have an expert interact with the data and choose this parameter, although automated approaches to similar problems have been suggested [22,15]. 3.2 Warping with the PAA representation In Section 2 we showed how to perform dynamic time warping on two sequences Q and C. Here we will show how to perform dynamic time warping using the reduced dimensionality versions of Q and C, which we denote i Q and i C respectively. For clarity we ....

Keogh, E., Smyth, P. (1997). A probabilistic approach to fast pattern matching in time series databases. Proc. of the 3 rd Mining. pp 24-20, AAAI Press.

No context found.

E. J. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In Proceedings of the 3rd International Conference on Knowledge Discovery and Data Mining (KDD'97), 1997.

No context found.

E. Keogh, P. Smyth, "A probabilistic approach to fast pattern matching in time series databases," Proc. Intl. Conf. Knowledge Discovery and Data Mining, 1997.

No context found.

Keogh, E., Smyt,h P.: A probabilistic approach to fast pattern matching in time series databases. Proc. of KDD (1997) 24--30

No context found.

E. Keogh and P. Smyth, A Probabilistic Approach to Fast Pattern Matching in Time Series Databases, Proceedings of the 3rd International Conference of Knowledge Discovery and Data Mining (1997), 24--29.

No context found.

E. Keogh and P. Smyth. A Probabilistic Approach to Fast Pattern Matching in Time Series Databases. In Proceedings of the Third ACM-SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 24--30, Newport Beach, California, USA, August 1997.

No context found.

E. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In the third conference on Knowledge Discovery in Databases and Data Mining, 1997.

No context found.

E. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In the third conference on Knowledge Discovery in Databases and Data Mining, 1997.

No context found.

Keogh, E., Smyth, P.: A probabilistic approach to fast pattern matching in time series databases. In: Proceedings of the 3rd International Conference of Knowledge Discovery and Data Mining. (1997) 24--30

No context found.

E. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. In the third conference on Knowledge Discovery in Databases and Data Mining, 1997.

No context found.

Keogh, E., Smyth, P.: A probabilistic approach to fast pattern matching in time series databases. In: Proceedings of the 3rd International Conference of Knowledge Discovery and Data Mining. (1997) 24--30

No context found.

E. Keogh and P. Smyth, "A Probabilistic Approach to Fast Pattern Matching in Time Series Databases," Proc. Third Int'l Conf. Knowledge Discovery and Data Mining, 1997.

No context found.

Keogh, E. J. and Smyth, P. (1997). A probabilistic approach to fast pattern matching in time series databases. In Proc. of the 3rd Int. Conf. on Knowl. Discovery and Data Mining, pages 20--24.

No context found.

Keogh, E. and Smyth, P. A probabilistic approach to fast pattern matching in time series databases. In Third International Conference on Knowledge Discovery and Data Mining, Heckerman, D., Mannila, H., Pregibon, D., and Uthurusamy, R., editors, 24--30. AAAI Press, Menlo Park,CA., (1997).

No context found.

E. J. Keogh and P. Smyth. A probabilistic approach to fast pattern matching in time series databases. Proc. Int. Conf. on Knowledge Discovery and Datamining, 24-30, 1997.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC