Abstract
Searching for characteristic patterns in time series is a topic addressed for decades by the research community. Conventional subsequence matching techniques usually rely on the definition of a target template pattern and a searching method for detecting similar patterns. However, the intrinsic variability of time series introduces changes in patterns, either morphologically and temporally, making such techniques not as accurate as desired. Intending to improve segmentation performances, in this paper, we proposed a Mask-based Neural Network (NN) which is capable of extracting desired patterns of interest from long time series, without using any predefined template. The proposed NN has been validated, alongside a subsequence matching algorithm, in two datasets: clinical (electrocardiogram) and human activity (inertial sensors). Moreover, the reduced dimension of the data in the latter dataset led to the application of transfer learning and data augmentation techniques to reach model convergence. The results have shown the proposed model achieved better segmentation performances than the baseline one, in both domains, reaching average Precision and Recall scores of 99.0% and 97.5% (clinical domain), along with 77.0% and 71.4% (human activity domain), introducing Neural Networks and Transfer Learning as promising alternatives for pattern searching in time series.
Highlights
IntroductionConventional techniques usually consist of a defined reference template, characterizing the pattern desired to match, and a distance metric (e.g., Euclidean Distance—ED, Dynamic Time Warping—DTW, Time Alignment Measurement—TAM [4], among others) measuring the similarity of that template relative to the portion of a signal evaluated [5]
In order to increase the flexibility of pattern segmentation in time series, rendering the task less sensitive to the latter’s variable components, as well as less domain-oriented and conditioned to user parameter choices, in this paper, we propose a Deep Learning (DL) architecture that performs a point-by-point mask-based segmentation of time series
A new Deep Learning approach has been proposed to improve the segmentation of patterns in time series, aiming to increase the robustness of the matching process, flexibly handling natural variability issues of such signals
Summary
Conventional techniques usually consist of a defined reference template, characterizing the pattern desired to match, and a distance metric (e.g., Euclidean Distance—ED, Dynamic Time Warping—DTW, Time Alignment Measurement—TAM [4], among others) measuring the similarity of that template relative to the portion of a signal evaluated [5]. An illustration of such an approach is displayed, where a template window is slid along with a longer time series at the same time a distance metric is computed.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
