Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2012:2012:280-288.
doi: 10.1145/2339530.2339578.

Mining Recent Temporal Patterns for Event Detection in Multivariate Time Series Data

Iyad Batal  1 , Dmitriy Fradkin  2 , James Harrison  3 , Fabian Moerchen  4 , Milos Hauskrecht  5
Affiliations

Mining Recent Temporal Patterns for Event Detection in Multivariate Time Series Data

Iyad Batal et al. KDD. 2012.

Abstract

Improving the performance of classifiers using pattern mining techniques has been an active topic of data mining research. In this work we introduce the recent temporal pattern mining framework for finding predictive patterns for monitoring and event detection problems in complex multivariate time series data. This framework first converts time series into time-interval sequences of temporal abstractions. It then constructs more complex temporal patterns backwards in time using temporal operators. We apply our framework to health care data of 13,558 diabetic patients and show its benefits by efficiently finding useful patterns for detecting and diagnosing adverse medical conditions that are associated with diabetes.

Keywords: Event Detection; Patient Classification; Temporal Abstractions; Temporal Pattern Mining; Time-interval Patterns.

PubMed Disclaimer

Figures

Figure 1
Figure 1. An example of an EHR data instance with three temporal variables. The black dots represent their values over time
Figure 2
Figure 2. An MSS representing 24 days of a patient record. In this example, there are two temporal variables (creatinine and glucose)
Figure 3
Figure 3. Allen’s temporal relations
Figure 4
Figure 4. A temporal pattern with states 〈(C, H), (G, N), (B, H), (G, H)〉 and temporal relations R1,2 = c, R1,3 = c, R1,4 = b, R2,3 = c, R2,4 = b and R3,4 = c
Figure 5
Figure 5. The number of temporal patterns of TP and RTP on all major diagnosis datasets (minimum support is 15%)
Figure 6
Figure 6. The mining time (in seconds) of TP_Apriori, RTP_no-lists, TP_lists and RTP_lists on all major diagnosis datasets (minimum support is 15%)
Figure 7
Figure 7. The mining time (in seconds) of TP_Apriori, RTP_no-lists, TP_lists and RTP_lists on the CARDI dataset for different minimum support values
Figure 8
Figure 8. The mining time (in seconds) of TP_Apriori, RTP_no-list, TP_lists and RTP_lists on the CARDI dataset for different maximum gap values (in months). The minimum support is 15%
See this image and copyright information in PMC

References

    1. Agrawal R, Srikant R. Fast algorithms for mining association rules in large databases; Proceedings of VLDB; 1994.
    1. Agrawal R, Srikant R. Mining sequential patterns; Proceedings of ICDE; 1995.
    1. Allen F. Towards a general theory of action and time. Artificial Intelligence. 1984;23:123–154.
    1. Batal I, Hauskrecht M. A supervised time series feature extraction technique using dct and dwt; Proceedings of ICMLA; 2009.
    1. Batal I, Valizadegan H, Cooper GF, Hauskrecht M. A pattern mining approach for classifying multivariate temporal data; Proceedings of the IEEE international conference on bioinformatics and biomedicine (BIBM); 2011. - PMC - PubMed

LinkOut - more resources