Attention to Warp: Deep Metric Learning for Multivariate Time Series

Matsuo, Shinnosuke; Wu, Xiaomeng; Atarsaikhan, Gantugs; Kimura, Atsushi; Kashino, Kunio; Iwana, Brian Kenji; Uchida, Seiichi

doi:10.1007/978-3-030-86334-0_23

Cited by 3 publications

(2 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, there is a preliminary conference paper of this work [28] and this paper contains significant differences from it. First, we newly propose a plug-in scenario, where our deep attentive time warping is utilized as a differentiable module in a large classification system.…”

Section: Dtw With Deep Metric Learningmentioning

confidence: 99%

“…We further confirmed that the plug-in usage of our technique achieves the state-of-the-art performance in large-scale signature verification tasks. Moreover, for the stand-alone scenario, we conduct more extensive comparison experiments on over 50 classification tasks in UCR dataset, whereas only four tasks have been tackled in [28]. Technical details are also newly elaborated in this paper.…”

Section: Dtw With Deep Metric Learningmentioning

confidence: 99%

See 1 more Smart Citation

Deep attentive time warping

Matsuo

Atarsaikhan

et al. 2023

Pattern Recognition

View full text Add to dashboard Cite

Section: Dtw With Deep Metric Learningmentioning

confidence: 99%

Section: Dtw With Deep Metric Learningmentioning

confidence: 99%

Deep attentive time warping

Matsuo

Atarsaikhan

et al. 2023

Pattern Recognition

View full text Add to dashboard Cite

Towards Backdoor Attack on Deep Learning based Time Series Classification

Ding

Zhang

Huang

et al. 2022

2022 IEEE 38th International Conference on Data Engineering (ICDE)

View full text Add to dashboard Cite

Finding Representative Sampling Subsets in Sensor Graphs Using Time-series Similarities

Chakraborty

Holm

Pedersen

et al. 2023

ACM Trans. Sen. Netw.

View full text Add to dashboard Cite

With the increasing use of IoT-enabled sensors, it is important to have effective methods to query the sensors. For example, in a dense network of battery-driven temperature sensors, it is often possible to query (sample) only a subset of the sensors at any given time, since the values of the non-sampled sensors can be estimated from the sampled values. If we can divide the set of sensors into disjoint so-called representative sampling subsets that each represents all the other sensors sufficiently well, we can alternate between the sampling subsets and thus, increase the battery life significantly of the sensor network. In this paper, we formulate the problem of finding representative sampling subsets as a graph problem on a so-called sensor graph with the sensors as nodes. Our proposed solution, SubGraphSample , consists of two phases. In Phase-I, we create edges in the similarity graph based on the similarities between the time-series of sensor values, analyzing six different techniques based on proven time-series similarity metrics. In Phase-II, we propose six different sampling techniques to find the maximum number of representative sampling subsets . Finally, we propose AutoSubGraphSample which auto-selects the best technique for Phase-I and Phase-II for a given dataset. Our extensive experimental evaluation shows that AutoSubGraphSample can yield significant battery life improvements within realistic error bounds.

show abstract

Attention to Warp: Deep Metric Learning for Multivariate Time Series

Cited by 3 publications

References 33 publications

Deep attentive time warping

Deep attentive time warping

Towards Backdoor Attack on Deep Learning based Time Series Classification

Finding Representative Sampling Subsets in Sensor Graphs Using Time-series Similarities

Contact Info

Product

Resources

About