Enhancing performance and interpretability of multivariate time-series model through sparse saliency

Simple item page
dc.contributor.authorKong, Xiangqi
dc.contributor.authorXing, Yang
dc.contributor.authorLiu, Zeyu
dc.contributor.authorTsourdos, Antonios
dc.contributor.authorWikander, Andreas
dc.date.accessioned2025-04-14T10:25:31Z
dc.date.available2025-04-14T10:25:31Z
dc.date.freetoread2025-04-14
dc.date.issued2024-09-24
dc.date.pubOnline2025-03-20
dc.description.abstractExplainable time-series modelling is an essential task for modern intelligent transportation systems (ITS). How-ever, balancing accuracy and interpretability in multivariate time series forecasting presents significant challenges. These challenges arise from the necessity to understand the significance of features and their temporal variations. Factors such as autocorrelation in time series and data processing techniques like sliding windows expand feature sets, thereby complicating pattern recognition using traditional post-hoc explanation methods and making the issue even more complex. To overcome these challenges, in this study, we propose a flexible post-process approach which generates sparse and normalized saliency values based on existing saliency generation methods such as GradientSHAP. Additionally, an optional window aggregation and alignment strategy is introduced to align with the original time series dataset, enhancing the intuitive understanding of feature importance. Furthermore, the potential use of sparse saliency for data augmentation to improve the model is explored. Lastly, we utilize naturalistic data from San Francisco airport to demonstrate our approach for ITS time-series prediction and explanation. The evaluation results indicate that integrating sparse saliency from high-performing models not only boosts the performance of XGBoost models by 10.92% but also simplifies model complexity, facilitating easier interpretation.
dc.description.conferencename2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC)
dc.description.sponsorshipEngineering and Physical Sciences Research Council (EPSRC)
dc.description.sponsorshipThis project is supported by the Engineering and Physical Sciences Research Council (EPSRC) training grant entitled “DTP 2020-2021 Cranfield University” bearing reference EP/T518104/1.
dc.format.extentpp. 3265-3270
dc.identifier.citationKong X, Xing Y, Liu Z, et al., (2024) Enhancing performance and interpretability of multivariate time-series model through sparse saliency. In: 2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), 24-27 September 2024, Edmonton, Canada, pp. 3265-3270
dc.identifier.eissn2153-0017
dc.identifier.elementsID566980
dc.identifier.issn2153-0009
dc.identifier.urihttps://doi.org/10.1109/itsc58415.2024.10920008
dc.identifier.urihttps://dspace.lib.cranfield.ac.uk/handle/1826/23759
dc.language.isoen
dc.publisherIEEE
dc.publisher.urihttps://ieeexplore.ieee.org/document/10920008
dc.rightsAttribution 4.0 Internationalen
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/
dc.subject4605 Data Management and Data Science
dc.subject46 Information and Computing Sciences
dc.subjectGeneric health relevance
dc.titleEnhancing performance and interpretability of multivariate time-series model through sparse saliency
dc.typeConference paper
dcterms.coverageEdmonton, Canada
dcterms.dateAccepted2024-07-25
dcterms.temporal.endDate27 Sep 2024
dcterms.temporal.startDate24 Sep 2024

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Enhancing_performance_and_interpretability-2025.pdf
Size:
1014.38 KB
Format:
Adobe Portable Document Format
Description:
Accepted version
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.63 KB
Format:
Plain Text
Description:
Download

Collections

Staff publications (AIRS)