Experiencing SAX: a novel symbolic representation of time series

Lin, Jessica; Keogh, Eamonn; Li, Wei; Lonardi, Stefano

doi:10.1007/s10618-007-0064-z

Cited by 1,355 publications

(940 citation statements)

References 42 publications

Supporting

Mentioning

851

Contrasting

Unclassified

Order By: Relevance

“…In the wide field of time-series research, in particular regarding data management and data mining, several techniques have been proposed that can be used to generate an abstract representation of time series [5,10,18,20,24,38,39,43]. This includes Fourier transforms, wavelets, symbolic representations and piecewise regression.…”

Section: Time-series Representationsmentioning

confidence: 99%

See 1 more Smart Citation

A time-series compression technique and its application to the smart grid

Eichinger

Efros

Karnouskos³

et al. 2014

The VLDB Journal

View full text Add to dashboard Cite

Time-series data is increasingly collected in many domains. One example is the smart electricity infrastructure, which generates huge volumes of such data from sources such as smart electricity meters. Although today this data is used for visualization and billing in mostly 15-min resolution, its original temporal resolution frequently is more finegrained, e.g., seconds. This is useful for various analytical applications such as short-term forecasting, disaggregation and visualization. However, transmitting and storing huge amounts of such fine-grained data is prohibitively expensive in terms of storage space in many cases. In this article, we present a compression technique based on piecewise regression and two methods which describe the performance of the compression. Although our technique is a general approach for time-series compression, smart grids serve as our running example and as our evaluation scenario. Depending on the data and the use-case scenario, the technique compresses data by ratios of up to factor 5,000 while maintaining its usefulness for analytics. The proposed technique has outperformed related work and has been applied to three real-world energy datasets in different scenarios. Finally, we show that the proposed compression technique can be implemented in a state-of-the-art database management system.

show abstract

Section: Time-series Representationsmentioning

confidence: 99%

“…Another approach for time-series compression is symbolic data representation, which is based on discretization [24,39]. This has recently been applied to smart-meter data [43].…”

Section: Time-series Representationsmentioning

confidence: 99%

A time-series compression technique and its application to the smart grid

Eichinger

Efros

Karnouskos³

et al. 2014

The VLDB Journal

View full text Add to dashboard Cite

show abstract

“…Table 1 shows the major techniques arranged in a hierarchy. (Kumar et al 2005) Data adaptive Piecewise Polynomials Interpolation* (Morinaka et al 2001) Regression (Shatkay and Zdonik 1996) Adaptive Piecewise Constant Approximation* (Keogh et al 2001b) Singular Value Decomposition* Symbolic Natural Language (Portet et al 2007) Strings (Huang and Yu 1999) Non-Lower Bounding (André-Jönsson and Badal 1997; Huang and Yu 1999;Megalooikonomou et al 2005) SAX* (Lin et al 2007), iSAX*…”

Section: Time Series Representationsmentioning

confidence: 99%

“…Our approach is based on a modification of the SAX representation to allow extensible hashing (Lin et al 2007). That is, the number of bits used for evaluation of our representation can be dynamically changed, corresponding to a desired resolution.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

iSAX: disk-aware mining and indexing of massive time series datasets

Shieh

Keogh

2009

Data Min Knowl Disc

Self Cite

View full text Add to dashboard Cite

Current research in indexing and mining time series data has produced many interesting algorithms and representations. However, the algorithms and the size of data considered have generally not been representative of the increasingly massive datasets encountered in science, engineering, and business domains. In this work, we introduce a novel multi-resolution symbolic representation which can be used to index datasets which are several orders of magnitude larger than anything else considered in the literature. To demonstrate the utility of this representation, we constructed a simple tree-based index structure which facilitates fast exact search and orders of magnitude faster, approximate search. For example, with a database of one-hundred million time series, the approximate search can retrieve high quality nearest neighbors in slightly over a second, whereas a sequential scan would take tens of minutes. Our experimental evaluation demonstrates that our representation allows index performance to scale well with increasing dataset sizes. Additionally, we provide analysis concerning parameter sensitivity, approximate search effectiveness, and lower bound comparisons between time series representations in a bit constrained environment. We further show how to exploit the combination of both exact and approximate search as sub-routines in data mining algorithms, allowing for the exact mining of truly massive real world datasets, containing tens of millions of time series.Responsible editor: Bart Goethals.

show abstract

Anomaly Detection in Sequential Data: Principles and Case Studies

Andrade

Gama

Ribeiro

et al. 2019

Wiley Encyclopedia of Electrical and Electronics Engineering

View full text Add to dashboard Cite

Humans are frequently looking for patterns and uniformity to support their choices and decisions. Whatever falls outside the expected can be said to be an anomaly. However, in many practical situations, the presence of anomalies can provide valuable insights, which can point out useful novelties. Thus, in predictive maintenance, for example, anomaly detection is useful to predict equipment failures and prevent losses at technical and financial levels. Most modern equipment has logging systems that allow to collect a high diversity of data regarding the equipment operation and working conditions. One such case is about several human activities. In both applications, looking for anomalies is a relevant task. The large amount of data, collected during a long period of time, makes the analysis by humans unfeasible. Data mining techniques can automatically extract models for anomaly and novelty detection from these data. These techniques, when used in predictive systems, are able to detect anomalies and issue an alert before they start, avoiding interruptions and breakdowns. After briefly describing the main aspects of anomaly detection and the most popular approaches and techniques, this article presents an overview of the main anomaly detection techniques used for predictive maintenance.

show abstract

Experiencing SAX: a novel symbolic representation of time series

Cited by 1,355 publications

References 42 publications

A time-series compression technique and its application to the smart grid

A time-series compression technique and its application to the smart grid

iSAX: disk-aware mining and indexing of massive time series datasets

Anomaly Detection in Sequential Data: Principles and Case Studies

Contact Info

Product

Resources

About