English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Classification of multi-dimensional time series from real-world systemsrequire fine-grained learning of complex features such as cross-dimensionaldependencies and intra-class variations-all under the practical challenge oflow training data availability. However, standard deep learning (DL) strugglesto learn generalizable features in low-data environments due to modeloverfitting. We propose a versatile yet data-efficient framework, IntelligentlyAugmented Contrastive Tensor Factorization (ITA-CTF), to learn effectiverepresentations from multi-dimensional time series. The CTF module learns coreexplanatory components of the time series (e.g., sensor factors, temporalfactors), and importantly, their joint dependencies. Notably, unlike standardtensor factorization (TF), the CTF module incorporates a new contrastive lossoptimization to induce similarity learning and class-awareness into the learntrepresentations for better classification performance. To strengthen thiscontrastive learning, the preceding ITA module generates targeted butinformative augmentations that highlight realistic intra-class patterns in theoriginal data, while preserving class-wise properties. This is achieved bydynamically sampling a "soft" class prototype to guide the warping of eachquery data sample, which results in an augmentation that is intelligentlypattern-mixed between the "soft" class prototype and the query sample. Theseaugmentations enable the CTF module to recognize complex intra-class variationsdespite the limited original training data, and seek out invariant class-wiseproperties for accurate classification performance. The proposed method iscomprehensively evaluated on five different classification tasks. Compared tostandard TF and several DL benchmarks, notable performance improvements up to18.7% were achieved.

Intelligently Augmented Contrastive Tensor Factorization: Empowering Multi-dimensional Time Series Classification in Low-Data Environments