Real-Time Classification of Streaming Sensor Data

Kasetty, Shruti; Stafford, Candice A.; Walker, G. P.; Wang, Xiaoyue; Keogh, Eamonn

doi:10.1109/ictai.2008.143

Cited by 26 publications

(17 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Kasetty et al (2008) demonstrate that their SAX-based approach to mining time series data does not have this issue. Other future work includes applying it to domains with an even wider variety of sampling rates in order to fully observe the effects of the parameters.…”

Section: Discussion and Future Workmentioning

confidence: 96%

“…There are many other time series analysis approaches that address related problems (e.g. Tanaka and Uehara 2003;Yin and Gaber 2008;Cheng and Tan 2008;Kasetty et al 2008) but none of these address the issue of multi-dimensional motif discovery with the additional task of identifying the most relevant dimensions of the data. Recent work on indexable SAX (Shieh and Keogh 2009) has introduced a very efficient way to store and search large time series and this work could make our approach even more efficient.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Identifying predictive multi-dimensional time series motifs: an application to severe weather prediction

McGovern

Rosendahl

Brown

et al. 2010

Data Min Knowl Disc

View full text Add to dashboard Cite

We introduce an efficient approach to mining multi-dimensional temporal streams of real-world data for ordered temporal motifs that can be used for prediction. Since many of the dimensions of the data are known or suspected to be irrelevant, our approach first identifies the salient dimensions of the data, then the key temporal motifs within each dimension, and finally the temporal ordering of the motifs necessary for prediction. For the prediction element, the data are assumed to be labeled. We tested the approach on two real-world data sets. To verify the generality of the approach, we validated the application on several subjects from the CMU Motion Capture database. Our main application uses several hundred numerically simulated supercell thunderstorms where the goal is to identify the most important features and feature interrelationships which herald the development of strong rotation in the lowest altitudes of a storm. We identified sets of precursors, in the form of meteorological Responsible editor:123 Identifying predictive multi-dimensional time series motifs 233 quantities reaching extreme values in a particular temporal sequence, unique to storms producing strong low-altitude rotation. The eventual goal is to use this knowledge for future severe weather detection and prediction algorithms.

show abstract

Section: Discussion and Future Workmentioning

confidence: 96%

mentioning

confidence: 99%

Identifying predictive multi-dimensional time series motifs: an application to severe weather prediction

McGovern

Rosendahl

Brown

et al. 2010

Data Min Knowl Disc

View full text Add to dashboard Cite

show abstract

“…Some more recent publications explicitly focus on the real-time analysis and visualization aspects of time series: The work of Kasetty et al [11] focuses on realtime classification of streaming sensor data. The authors use Symbolic Aggregate Approximation (SAX) [15] to transform the time series data and then show how time series bitmaps representing pattern frequencies can be updated in constant time for classifying high-rate data streams.…”

Section: Time Series Visualizationmentioning

confidence: 99%

Dynamic Visual Analytics—Facing the Real-Time Challenge

Mansmann

Fischer

Keim

2012

Expanding the Frontiers of Visual Analytics and Visualization

View full text Add to dashboard Cite

Abstract. Modern communication infrastructures enable more and more information to be available in real-time. While this has proven to be useful for very targeted pieces of information, the human capability to process larger quantities of mostly textual information is definitely limited. Dynamic visual analytics has the potential to circumvent this real-time information overload by combining incremental analysis algorithms and visualizations to facilitate data stream analysis and provide situational awareness. In this book chapter we will thus define dynamic visual analytics, discuss its key requirements and present a pipeline focusing on the integration of human analysts in real-time applications. To validate this pipeline, we will demonstrate its applicability in a real-time monitoring scenario of server logs.

show abstract

“…In the literature, a number of representation techniques have been developed for time series data, such as Symbolic Aggregate Approximation (SAX) [6], Discrete Fourier Transformation (DFT) [7], and Discrete Wavelet Transformation (DTW) [8]. Among them, SAX is one of the most popular representation techniques, which has proven its effectiveness in dimensionality reduction for time series data [9,10,11,12].…”

Section: Introductionmentioning

confidence: 99%

Time Series Representation: A Random Shifting Perspective

Bai

Xiong

Zhu

et al. 2013

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract.A long standing challenge for time series analysis is to develop representation techniques for dimension reduction while still preserving their fundamental features. As an effective representation technique, Symbolic Aggregate Approximation (SAX) has been widely used for dimension reduction in time series analysis. However, SAX always maps time series data into symbols by definite breakpoints. As a result, the similar points close to the breakpoints cannot be well represented, and thus lead to poor Tightness of Lower Bounds (TLB). To fill this crucial void, in this paper, we develop a time series representation method, named Random Shifting based SAX (rSAX), which has the ability in significantly improving the TLB of representations without increasing the corresponding granularity of representations. Specifically, the key idea of rSAX is to generate a group of breakpoints by random shifting rather than definite breakpoints. Therefore, the points close to each other will have higher probabilities to be mapped into the same symbols, while the points far away from each other will have higher probabilities to be mapped into different symbols. In addition, we also theoretically prove that rSAX can achieve better mapping performances and TLB than SAX. Finally, extensive experiments on several real-world data sets clearly validate the effectiveness and efficiency of the rSAX approach.

show abstract

Real-Time Classification of Streaming Sensor Data

Abstract: Abstract

Cited by 26 publications

References 21 publications

Identifying predictive multi-dimensional time series motifs: an application to severe weather prediction

Identifying predictive multi-dimensional time series motifs: an application to severe weather prediction

Dynamic Visual Analytics—Facing the Real-Time Challenge

Time Series Representation: A Random Shifting Perspective

Contact Info

Product

Resources

About