Korn, F., Jagadish, H. and Faloutsos, C. (1997): Efficiently supporting ad hoc queries in large datasets of time sequences. Proc. ACM SIGMOD International Conference on Management of Data, Tuescon, AZ, USA, 289-300.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....like clustering, classification and association rule mining. Time series databases convert time series segments to multidimensional points using some transformation (e.g. Discrete Fourier Transform (DFT) 5, 46] Discrete Wavelet Transform (DWT) 29, 79] Singular Value Decomposition (SVD) [79, 76, 81]) Similarity search is then performed on the transformed data. Example applications include a doctor searching for a particular pattern (that implies a heart irregularity) in the ECG database for diagnosis, a stock analyst searching for a particular pattern in the stock database for prediction ....

....the transformed space. The technique was introduced in [5] and extended in [119, 32] The original work by Agrawal et al. utilizes the Discrete Fourier Transform (DFT) to perform the dimensionality reduction, but other techniques have been suggested, including Singular Value Decomposition (SVD) [79, 76, 81], the Discrete Wavelet Transform (DWT) 29, 151, 75] and Piecewise Aggregate Approximation (PAA) 79, 153] For a given index structure, the efficiency of indexing depends only on the fidelity of the approximation in the reduced dimensionality space. However, in choosing a dimensionality ....

[Article contains additional citation context not shown here]

F. Korn, H. Jagadish, and C. Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequences. Proc. of SIGMOD, 1997. 163

....to construct an efficient grid structure in bounded DFT feature space. This is what enable us to give real time results. 19] allows false negatives, whereas our method does not. Other techniques such as Discrete Wavelet Trans form (DWT) 5, 26, 22, 12] Singular Value Decomposition (SVD)[18] and Piecewise Constant Approximation (PCA) 27, 17] are also proposed for similarity search. Keogh et al. 17] compares these techniques for time series similarity queries. The performance of these techniques varied depending on the characteristics of the datasets, because no single transform can ....

F. Korn, H. V. Jagadish, and C. Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequences. In Proceedings ACM SIGMOD International Conference on Management of Data, pages 289-300, 1997.

....the transformed space. The technique was introduced in [1] and extended in [39, 31,11] The original work by Agrawal et al. utilizes the Discrete Fourier Transform (DFT) to perform the dimensionality reduction, but other techniques have been suggested, including Singular Value Decomposition (SVD) [28, 24, 23], the Discrete Wavelet Transform (DWT) 9, 49, 22] and Piecewise Aggregate Approximation (PAA) 24, 52] For a given index structure, the efficiency of indexing depends only on the fidelity of the approximation in the reduced dimensionality space. However, in choosing a dimensionality reduction ....

....time series data. In [16] the authors introduced GEneric Multimedia INdexing method (GEMINI) which can exploit any dimensionality reduction method to allow efficient indexing. The technique was originally introduced for time series, but has been successfully extend to many other types of data [28]. An important result in [16] is that the authors proved that in order to guarantee no false dismissals, the distance measure in the index space must satisfy the following condition: Dindexspace(A,B) Dte(A,B) This theorem is known as the lower bounding lemma or the contractive property. ....

[Article contains additional citation context not shown here]

Korn, F., Jagadish, H & Faloutsos. C. (1997). Efficiently supporting ad hoc queries in large datasets of time sequences. Proceedings of SIGMOD '97, Tucson, AZ, pp 289-300.

.... Wavelet Based Anytime Algorithm for K Means Clustering of Time Series Michail Vlachos Jessica Lin Eamonn Keogh Dimitrios Gunopulos Computer Science Engineering Department University of California Riverside Riverside, CA 92521 mvlachos, jessica, eamonn, dg cs.ucr.edu ABSTRACT The emergence of the field of data mining in the last decade has sparked an increasing interest in clustering of tiate series. Although there has been much research on clustering in general, most classic machine learning and data mining ....

....for clustering large datasets of time series, k Means is preferable due to its faster running time. In order to scale the various clustering methods to massive datasets, one can either reduce the number of objects, N, by sampling [5] or reduce the dimensionality of the objects [1, 6, 14, 25, 29, 35, 36, 22, 23] In the case of time series, the objective is to find a representation at a lower dimensionality that preserves the original information and describes the original shape of the time series data as closely as possible. Many approaches have been suggested in the literature, ....

[Article contains additional citation context not shown here]

Korn, F., Jagadish, H. & Faloutsos, C. (1997). Efficiently supporting ad hoc queries in large datasets of time sequences. In proceedings of the ACM SIGMOD InFl Conference on Management of Data. Tucson, AZ, May 13-15. pp 289-300.

....weather data, DNA sequences, and sensor data from robotics. New emerging applications, such as data mining and information retrieval by content, require the capability of finding similar patterns, i.e. similarity query. Similarity query on persistent datasets has received a lot of attentions ([1, 21, 11, 15, 16, 20]) however to the best of our knowledge, there are no prior studies on pattern recognition isolation over CDS. Trivially, the performance of a similarity query is determined largely by the chosen distance metric. The most straightforward approach for measuring the similaritybetween two sequences ....

....with length as a pointin D space, and rotate the axes. This is exactly what singular value decomposition (SVD) does, but SVD does this in an optimal way (in terms of L 2 norm) for the given dataset# the reason is that effectively SVD maximizes the variance along the first few rotations [16]thus gives the optimal decomposition of the dataset byway of rotations. Furthermore, the nature of our data requires a 2 D transformation in case of DFT or DWT# however, since our datasets are not correlated on the sensor dimension at anygiven time, we do not expect DFT or DWT to perform well. ....

F. Korn, H. V. Jagadish, and C. Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequences. In J. Peckham, editor, SIGMOD 1997.

....GEMINI framework of Faloutsos [17] but suggest a different approach to the dimensionality reduction stage. The proposed representations include the Discrete Fourier Transform (DFT) 1, 11, 16, 28, 49, 50] several kinds of Wavelets (DWT) 10, 27, 45, 51, 57, 60] Singular Value Decomposition [32, 35], Adaptive Piecewise Constant Approximation [32] Inner Products [18] and Piecewise Aggregate Approximation (PAA) 61] The majority of work has focused solely on performance issues, however some authors have also considered other issues such as supporting non Euclidean distance measures [32, 50, ....

....Note that for both problems, informal experiments suggest humans can achieve an error rate of zero. For simplicity we use the 1 Nearest Neighbor algorithm, evaluated using leaving oneout . We compare the proposed methods to the simplest strawman, Euclidean distance. This measure is well known [1, 10, 11, 13, 14, 16, 17, 18, 27, 32, 35, 36, 40, 45, 49, 50, 60, 61, 62], parameterless, trivial to implement and predates data mining by several decades. We originally intended to implement every proposed similarity measure in our survey, but several of the papers do not include a detailed enough description to allow reimplementafion [39, 48] We contented ourselves ....

Kom, F., Jagadish, H. & Faloutsos, C. (1997). Efficiently supporting ad hoc queries in large datasets of time sequences. In proceedings of the ACM SIGMOD Int7 Conference on Management of Data. Tucson, AZ, May 13-15. pp 289-300.

....to construct an efficient grid structure in bounded DFT feature space. This is what enable us to give real time results. 19] allows false negatives, whereas our method does not. Other techniques such as Discrete Wavelet Trans form (DWT) 5, 26, 22, 12] Singular Value Decomposition (SVD)[18] and Piecewise Constant Approximation (PCA) 27, 17] are also proposed for similarity search. Keogh et al. 17] compares these techniques for time series similarity queries. The performance of these techniques varied depending on the characteristics of the datasets, because no single transform can ....

F. Korn, H. V. Jagadish, and C. Faloutsos. Effi- ciently supporting ad hoc queries in large datasets of time sequences. In Proceedings A CM SIGMOD International Conference on Management of Data, pages 289-300, 1997.

....space is restricted. Wu et. al [19] employ SVD (Singular Value Decomposition) and DFT (Discrete Fourier Transform) for reducing the dimension of feature vectors in the problem of searching images in large image databases. SVD and DFT have been widely used in time series databases as well [12, 1]. One of the main challenges with our application is that the amount of required space to store scientific observation data is large. Thus, we also need some compression scheme to reduce the size of the required disk space. SVD and DFT could be useful in this case. However, SVD and DFT lack the ....

F. Korn, H. V. Jagadish, andC. Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequences. Proceedings of the ACM SIGMOD international conference on Management of data, 26(2):289--300, 1997.

....transformed space. The technique was introduced in [1] and extended in [19, 6, 21, 10] The original work by Agrawal et al. utilizes the Discrete Fourier Transform (DFT) to perform the dimensionality reduction, but other techniques have been suggested, including Singular Value Decomposition (SVD) [18, 15], the Discrete Wavelet Transform (DWT) 5, 24] and Piecewise Polynomial Approximations [15, 26] In general, the efficiency of indexing depends only on the fidelity of the approximation in the reduced dimensionality space. Clearly no single dimensionality reduction technique can be optimal on all ....

....structures to index time series data. In [10] the authors introduced GEneric Multimedia INdexIng method (GEMINI) which can exploit any dimensionality reduction method to allow indexing. The technique was originally introduced for time series, but has been successfully extended to other data types [18]. An important result in [10] is that the authors proved that in order to guarantee no false dismissals, the distance measure in the index space must satisfy the following condition: D index space (A,B) D true (A,B) 2) This theorem is known as the lower bounding lemma or the contractive ....

[Article contains additional citation context not shown here]

Korn, F., Jagadish, H & Faloutsos. C. (1997). Efficiently supporting ad hoc queries in large datasets of time sequences. Proceedings of SIGMOD, Tucson, AZ, pp 289300.

....the transformed space. The technique was introduced in [1] and extended in [39, 31,11] The original work by Agrawal et al. utilizes the Discrete Fourier Transform (DFT) to perform the dimensionality reduction, but other techniques have been suggested, including Singular Value Decomposition (SVD) [28, 24, 23], the Discrete Wavelet Transform (DWT) 9, 49, 22] and Piecewise Aggregate Approximation (PAA) 24, 52] For a given index structure, the efficiency of indexing depends only on the fidelity of the approximation in the reduced dimensionality space. However, in choosing a dimensionality reduction ....

....time series data. In [16] the authors introduced GEneric Multimedia INdexIng method (GEMINI) which can exploit any dimensionality reduction method to allow efficient indexing. The technique was originally introduced for time series, but has been successfully extend to many other types of data [28]. An important result in [16] is that the authors proved that in order to guarantee no false dismissals, the distance measure in the index space must satisfy the following condition: D index space (A,B) D true (A,B) 2) This theorem is known as the lower bounding lemma or the contractive ....

[Article contains additional citation context not shown here]

Korn, F., Jagadish, H & Faloutsos. C. (1997). Efficiently supporting ad hoc queries in large datasets of time sequences. Proceedings of SIGMOD '97, Tucson, AZ, pp 289-300.

....the transformed space. The technique was introduced in [1] and extended in [39, 31,11] The original work by Agrawal et al. utilizes the Discrete Fourier Transform (DFT) to perform the dimensionality reduction, but other techniques have been suggested, including Singular Value Decomposition (SVD) [28, 24, 23], the Discrete Wavelet Transform (DWT) 9, 49, 22] and Piecewise Aggregate Approximation (PAA) 24, 52] For a given index structure, the efficiency of indexing depends only on the fidelity of the approximation in the reduced dimensionality space. However, in choosing a dimensionality reduction ....

....time series data. In [16] the authors introduced GEneric Multimedia INdexIng method (GEMINI) which can exploit any dimensionality reduction method to allow efficient indexing. The technique was originally introduced for time series, but has been successfully extend to many other types of data [28]. An important result in [16] is that the authors proved that in order to guarantee no false dismissals, the distance measure in the index space must satisfy the following condition: D index space (A,B) D true (A,B) 2) This theorem is known as the lower bounding lemma or the contractive ....

[Article contains additional citation context not shown here]

Korn, F., Jagadish, H & Faloutsos. C. (1997). Efficiently supporting ad hoc queries in large datasets of time sequences. Proceedings of SIGMOD '97, Tucson, AZ, pp 289-300.

....to reduce the size of OTSA tree further by dropping nodes and or coefficients with less energy. The resulting condensed OTSA tree, hence, becomes a competitor to other techniques discussed in the literature such as Discrete Fourier Transform (DFT) 2, 3] and Single Value Decomposition (SVD) [20]. Therefore, we conducted comprehensive experiments to compare our results with those techniques. Briefly, we outperformed SVD and DFT in both performance and accuracy for our application. This is because SVD performs poorly when N M and DFT cannot maintain the surprises due to only keeping the ....

....techniques are expensive when compared with the wavelet transform and it cannot capture multi level surprises. In this paper, we also try to provide an efficient way for placing TSA tree on disk. By reducing data volume, we can achieve less disk I O and efficiency for some aggregate queries. In [20], Korn et. al use singular value decomposition (SVD) transform to map a huge matrix into a much smaller one. By data compression, they can support ad hoc queries on large volume of data sets. However, there are some problems inherent in the SVD method. With SVD, any M Theta N matrix X can be ....

[Article contains additional citation context not shown here]

F. Korn, H. V. Jagadish, and C. Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequences. Proceedings of the ACM SIGMOD international conference on Management of data, 26(2):289--300, 1997.

....circularity, transparency, relative area, rightangleness, sharpness, complexity, directedness and straightness Is ACME s symbol sufficiently similar to any other trademark symbols to cause confusion Table 1. Examples of feature extraction functions from several domains. 1. The authors of [KJF97] use feature extraction to perform ad hoc queries on large datasets of time sequences. The data consists of customer calling patterns from AT T and is in the order of hundreds of gigabytes. Calling patterns are stored in a matrix where each element has a numeric value. The rows correspond to ....

Flip Korn, H. V. Jagadish, and Christos Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequences. In Proc. SIGMOD Conference, 1997. Available on the Web at ftp://olympos.cs. umd.edu/pub/TechReports/sigmod97.ps.

No context found.

F. Korn, H.V. Jagadish, C. Faloutsos. Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. Proc. 1997 ACM-SIGMOD, pp. 289--300.

....more dramatic for multivariate data. The bulk of the database literature on histograms is focused towards query optimizers, as is our own work. However, approximate query answering is becoming increasingly important to provide data analysts with interactive responses from large data warehouses [6, 11]. To the extent that many data values (e.g. money amounts) stored in a data warehouse are drawn from a naturally continuous domain, the techniques we present in this paper are applicable to data warehousing contexts as well as to query optimizers. The paper is organized as follows: Section 2 ....

Flip Korn, H.V. Jagadish, and Christos Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequ ences. In Proc. ACM SIGMOD, pages 289--300, Tucson, AZ, May 1997.

No context found.

Korn, F., Jagadish, H. and Faloutsos, C. (1997): Efficiently supporting ad hoc queries in large datasets of time sequences. Proc. ACM SIGMOD International Conference on Management of Data, Tuescon, AZ, USA, 289-300.

No context found.

Korn F., Jagadish H. V., Faloutsos C.: Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. SIGMOD'97, AZ, USA, May 1997

No context found.

Korn, F., Jagadish, H. & Faloutsos, C. (1997). Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. In proceedings of the ACM SIGMOD Int'l Conference on Management of Data. Tucson, AZ, May 13-15. pp 289-300.

No context found.

Korn, F., Jagadish, H. & Faloutsos, C. (1997). Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. In proceedings of the ACM SIGMOD Int'l Conference on Management of Data. Tucson, AZ, May 13-15. pp 289-300.

No context found.

F. Korn, H. V. Jagadish, and C. Faloutsos. Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. In Proceedings of the 1997.

No context found.

F. Kom, H.V. Jagadish, and C. Faloutsos. Efficiently supporting ad hoc queries in large datasets of time sequences. In Proceedings of the ACM-SIGMOD International Conrence on Management of Data, Tucson, USA, 1997.

No context found.

F. Korn, H. V. Jagadish, and C. Faloutsos. Efficiently supporting Ad Hoc queries in large datasets of time sequences. Proc. ACM Conf. on Management of Data (SIGMOD), 289-300, 1997.

No context found.

Korn F., Jagadish H. V., Faloutsos C.: Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. SIGMOD'97, AZ, USA, May 1997

No context found.

Korn F., Jagadish H. V., Faloutsos C.: Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. SIGMOD'97, AZ, USA, May 1997

No context found.

Korn F., Jagadish H. V., Faloutsos C.: Efficiently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. SIGMOD'97, AZ, USA, May 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