Optimal Transport and Adaptive Thresholding for Universal Domain Adaptation on Time Series

2025-03-14

AI Summary

This paper addresses the challenges of unsupervised universal domain adaptation (UniDA) for time series, where source and target domains share only partial class overlap and target labels are entirely unavailable. To this end, we propose a robust knowledge transfer framework. Our key contributions are: (1) explicitly incorporating unknown-class samples into the optimal transport cost function to enhance discriminative distribution alignment; (2) constructing a joint decision space to improve inter-class separability; (3) designing a parameter-free adaptive thresholding mechanism to mitigate overconfident misclassification of unknown instances; and (4) integrating a Fourier-transform-based neural layer to strengthen temporal modeling capacity. Evaluated on multiple time-series UniDA benchmarks, our method achieves state-of-the-art performance, significantly improving both unknown-class detection accuracy and cross-domain generalization robustness.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Computer Vision: Adversarial Attacks & Robustness

Application Category

• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

Universal Domain Adaptation (UniDA) aims to transfer knowledge from a labeled source domain to an unlabeled target domain, even when their classes are not fully shared. Few dedicated UniDA methods exist for Time Series (TS), which remains a challenging case. In general, UniDA approaches align common class samples and detect unknown target samples from emerging classes. Such detection often results from thresholding a discriminability metric. The threshold value is typically either a fine-tuned hyperparameter or a fixed value, which limits the ability of the model to adapt to new data. Furthermore, discriminability metrics exhibit overconfidence for unknown samples, leading to misclassifications. This paper introduces UniJDOT, an optimal-transport-based method that accounts for the unknown target samples in the transport cost. Our method also proposes a joint decision space to improve the discriminability of the detection module. In addition, we use an auto-thresholding algorithm to reduce the dependence on fixed or fine-tuned thresholds. Finally, we rely on a Fourier transform-based layer inspired by the Fourier Neural Operator for better TS representation. Experiments on TS benchmarks demonstrate the discriminability, robustness, and state-of-the-art performance of UniJDOT.

Problem

• Addresses Universal Domain Adaptation for Time Series data.
• Improves detection of unknown target samples in domain adaptation.
• Reduces reliance on fixed or fine-tuned threshold values.

Innovation

• Optimal transport for unknown sample handling
• Auto-thresholding to reduce fixed threshold reliance
• Fourier transform layer for time series representation