🤖 AI Summary
To address insufficient modeling of translation equivariance and time-reversal invariance in multivariate time series classification (TSC), this paper proposes a dual-domain (time-frequency) multi-view modeling framework. It integrates frequency-domain spectral features extracted via continuous wavelet transform (CWT) with time-domain representations captured by temporal convolutions or MLPs. Crucially, it introduces the Mamba state-space model—the first such application in TSC—enabling linear-complexity modeling of long sequences. Additionally, a novel “tango” scanning mechanism is designed to enhance joint local-global sequence dependency modeling. Evaluated on ten standard benchmark datasets, the method achieves an average accuracy improvement of 6.45% over state-of-the-art approaches. This work establishes a new paradigm for multivariate TSC that balances theoretical soundness—through principled incorporation of equivariance and invariance properties—with computational efficiency.
📝 Abstract
Time series classification (TSC) on multivariate time series is a critical problem. We propose a novel multi-view approach integrating frequency-domain and time-domain features to provide complementary contexts for TSC. Our method fuses continuous wavelet transform spectral features with temporal convolutional or multilayer perceptron features. We leverage the Mamba state space model for efficient and scalable sequence modeling. We also introduce a novel tango scanning scheme to better model sequence relationships. Experiments on 10 standard benchmark datasets demonstrate our approach achieves an average 6.45% accuracy improvement over state-of-the-art TSC models.