Attention:The NSF Public Access Repository (PAR) system and access will be unavailable from 10:00 PM ET on Thursday, July 16 until 12:00 AM ET on Friday 17 due to maintenance. We apologize for the inconvenience.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: A New Attention Mechanism to Classify Multivariate Time Series
Classifying multivariate time series (MTS), which record the values of multiple variables over a continuous period of time, has gained a lot of attention. However, existing techniques suffer from two major issues. First, the long-range dependencies of the time-series sequences are not well captured. Second, the interactions of multiple variables are generally not represented in features. To address these aforementioned issues, we propose a novel Cross Attention Stabilized Fully Convolutional Neural Network (CA-SFCN) to classify MTS data. First, we introduce a temporal attention mechanism to extract long- and short-term memories across all time steps. Second, variable attention is designed to select relevant variables at each time step. CA-SFCN is compared with 16 approaches using 14 different MTS datasets. The extensive experimental results show that the CA-SFCN outperforms state-of-the-art classification methods, and the cross attention mechanism achieves better performance than other attention mechanisms.  more » « less
Award ID(s):
1757207 1633330 1914635
PAR ID:
10177556
Author(s) / Creator(s):
;
Date Published:
Journal Name:
Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence
Page Range / eLocation ID:
1999 to 2005
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; (, IEEE Transactions on Knowledge and Data Engineering)
    Multivariate time series (MTS) are collected for different variables in studying scientific phenomena or monitoring system health where one time series records the values of one variable for a time period. Among the different variables, it is common that only a few variables contribute significantly to a specific phenomenon. Furthermore, the variables contributing significantly to different phenomena are often different. We denote the different variables that contribute to the occurrences of different phenomena as Phenomenon-specific Variables (PVs). In this paper, we formulate a novel problem of identifying significant PVs from MTS datasets. To analyze MTS data, feature extraction techniques have been extensively studied. However, most of them identify important global features for one dataset and do not utilize the temporal order of time series. To solve the newly introduced problem, we propose a solution framework, CNNmts-X, which is a new variant of the Convolutional Neural Networks (CNN) and can embed other feature extraction techniques (as X). Furthermore, we design a CNNmts-LR method that implements a new feature identification approach(LR) as X in the CNNmts-X framework. The LR method leverages both Linear Discriminant Analysis (LDA) and Random Forest (RF). Our extensive experiments on five real datasets show that the CNNmts-LR method has exhibited much better performance than several other baseline methods. Using 30% of the PVs discovered from the CNNmts-LR, classifications can achieve better or simila performance than using all the variables. 
    more » « less
  2. ; ; ; (, 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC))
    Understanding neural function often requires multiple modalities of data, including electrophysiogical data, imaging techniques, and demographic surveys. In this paper, we introduce a novel neurophysiological model to tackle major challenges in modeling multimodal data. First, we avoid non-alignment issues between raw signals and extracted, frequency-domain features by addressing the issue of variable sampling rates. Second, we encode modalities through “cross-attention” with other modalities. Lastly, we utilize properties of our parent transformer architecture to model long-range dependencies between segments across modalities and assess intermediary weights to better understand how source signals affect prediction. We apply our Multimodal Neurophysiological Transformer (MNT) to predict valence and arousal in an existing open-source dataset. Experiments on non-aligned multimodal time-series show that our model performs similarly and, in some cases, outperforms existing methods in classification tasks. In addition, qualitative analysis suggests that MNT is able to model neural influences on autonomic activity in predicting arousal. Our architecture has the potential to be fine-tuned to a variety of downstream tasks, including for BCI systems. 
    more » « less
  3. ; ; ; ; ; ; ; ; ; ; et al (, Proceedings of the AAAI Conference on Artificial Intelligence)
    Adverse clinical events related to unsafe care are among the top ten causes of death in the U.S. Accurate modeling and prediction of clinical events from electronic health records (EHRs) play a crucial role in patient safety enhancement. An example is modeling de facto care pathways that characterize common step-by-step plans for treatment or care. However, clinical event data pose several unique challenges, including the irregularity of time intervals between consecutive events, the existence of cycles, periodicity, multi-scale event interactions, and the high computational costs associated with long event sequences. Existing neural temporal point processes (TPPs) methods do not effectively capture the multi-scale nature of event interactions, which is common in many real-world clinical applications. To address these issues, we propose the cross-temporal-scale transformer (XTSFormer), specifically designed for irregularly timed event data. Our model consists of two vital components: a novel Feature-based Cycle-aware Time Positional Encoding (FCPE) that adeptly captures the cyclical nature of time, and a hierarchical multi-scale temporal attention mechanism, where different temporal scales are determined by a bottom-up clustering approach. Extensive experiments on several real-world EHR datasets show that our XTSFormer outperforms multiple baseline methods. 
    more » « less
  4. ; (, Engineering with Computers)
    Abstract Multi-time-step (MTS) methods solve temporal multi-scale problems by decomposing the spatial problem domain into multiple subdomains and integrating each subdomain with a time step that is suitable for capturing its response. A critical limitation of existing MTS methods, however, is that they are based on enforcing continuity of velocities across the interface between subdomains and consequently lead to erroneous drifts in displacements that show up as nonphysical gaps and overlaps in the solution. In this article, we propose a new MTS framework that enables simultaneous enforcement of multiple continuity constraints between subdomains and the use of general interpolation schemes for associated Lagrange multipliers. To make the framework as general as possible, we first express almost all existing time integration schemes in a Generalized Runge–Kutta–Nyström (GRKN) form, which then serve as the underlying time integration schemes for individual subdomains. MTS methods designed using the proposed framework are able to achieve unconditional stability by introducing a scalar auxiliary variable (SAV) based on physical energy of the system and ensuring that this energy remains bounded during the simulation. We design new MTS methods based on piecewise-linear and quadratic interpolations of Lagrange multipliers with simultaneous enforcement of velocity and displacement continuity (VCDC) to demonstrate typical use-cases of the proposed framework. We present several numerical examples to show that solutions obtained using the new VCDC MTS methods are able to achieve greater accuracy for the same computational cost in comparison to single time integration schemes and existing MTS methods. 
    more » « less
  5. ; ; ; ; (, Proceedings)
    Accurate long-term predictions are the foundations for many machine learning applications and decision-making processes. Traditional time series approaches for prediction often focus on either autoregressive modeling, which relies solely on past observations of the target “endogenous variables”, or forward modeling, which considers only current covariate drivers “exogenous variables”. However, effectively integrating past endogenous and past exogenous with current exogenous variables remains a significant challenge. In this paper, we propose ExoTST, a novel transformer-based framework that effectively incorporates current exogenous variables alongside past context for improved time series prediction. To integrate exogenous information efficiently, ExoTST leverages the strengths of attention mechanisms and introduces a novel cross-temporal modality fusion module. This module enables the model to jointly learn from both past and current exogenous series, treating them as distinct modalities. By considering these series separately, ExoTST provides robustness and flexibility in handling data uncertainties that arise from the inherent distribution shift between historical and current exogenous variables. Extensive experiments on real-world carbon flux datasets and time series benchmarks demonstrate ExoTST's superior performance compared to state-of-the-art baselines, with improvements of up to 10% in prediction accuracy. Moreover, ExoTST exhibits strong robustness against missing values and noise in exogenous drivers, maintaining consistent performance in real-world situations where these imperfections are common. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Text Encoded Conversion
  • XML
  • Save / Share this Record
  • Send to Email