Normalized Cross-Match: Pattern Discovery Algorithm from Biofeedback Signals
Biofeedback signals are important elements in critical care applications, such as monitoring ECG data of a patient, discovering patterns from large amount of ECG data sets, detecting outliers from ECG data, etc. Because the signal data update continuously
- PDF / 759,417 Bytes
- 12 Pages / 439.37 x 666.142 pts Page_size
- 46 Downloads / 174 Views
3
Department of Computer and Information Science, University of Macau, Macau, China {yb47453,ccfong,fstasp,robertb}@umac.mo 2 School of Computer Science and Engineering, University of New South Wales, Sydney, Australia [email protected] Department of Computer Science, Electrical and Space Engineering, Lulea University of Technology, Lulea, Sweden [email protected]
Abstract. Biofeedback signals are important elements in critical care applications, such as monitoring ECG data of a patient, discovering patterns from large amount of ECG data sets, detecting outliers from ECG data, etc. Because the signal data update continuously and the sampling rates may be different, time-series data stream is harder to be dealt with compared to traditional historical time-series data. For the pattern discovery problem on time-series streams, Toyoda proposed the CrossMatch (CM) approach to discover the patterns between two timeseries data streams (sequences), which requires only O(n) time per data update, where n is the length of one sequence. CM, however, does not support normalization, which is required for some kinds of sequences (e.g. EEG data, ECG data). Therefore, we propose a normalized-CrossMatch approach (NCM) that extends CM to enforce normalization while maintaining the same performance capabilities.
Keywords: Pattern discovery streams
1
·
CrossMatch
·
NCM
·
Time-series
Introduction
Biofeedback signals in the form of time-series data are significant to critical care applications. For example, monitoring electrocardiogram (ECG) data of patient (ECG data is a kind of time-series data stream) helps to save human resources to monitor the ECG of the patient, discovering common patterns from large amount of ECG data sets may find specific pattern which represents the body condition of the patient, detecting outliers from ECG data helps to discover the problem of human health, etc. In general, most healthcare applications would have to deal with time-series, for analysis, matching, comparison and so forth. Other applications include but not limit to financial data, web click-stream data, c Springer International Publishing Switzerland 2016 H. Cao et al. (Eds.): PAKDD 2016 Workshops, LNAI 9794, pp. 169–180, 2016. DOI: 10.1007/978-3-319-42996-0 14
170
X. Gong et al.
motion capture data, and sensor network data. A number of different approaches for dealing with data streams have been proposed in the literature, e.g. subsequence matching [6,16], pattern discovery [17–19], and outlier detection [3,4].
(a) Sequence S
(b) Sequence T
Fig. 1. Discovered patterns in time-series data streams (Color figure online)
The objective of pattern discovery in data streams is to find common patterns among n data streams. Hence, the basis of the problem is to find common patterns between two data streams. Figure 1 shows an example of pattern discovery between two data streams. The red lines in Fig. 1 represent the patterns discovered in two respective sequences. Toyoda et al. [19] proposed an approach called CrossMatch (CM) t
Data Loading...
