Search papers, labs, and topics across Lattice.
This paper introduces MSRGC-Net, a novel framework for time-series clustering that leverages multiscale reservoir computing and granular-ball anchoring graph optimization to enhance both clustering effectiveness and computational efficiency. By utilizing a training-free approach, MSRGC-Net extracts multiscale temporal representations from raw time series data without the need for backpropagation, thus significantly lowering computational demands. Experimental results on various benchmark datasets show that MSRGC-Net outperforms existing state-of-the-art clustering methods while maintaining superior efficiency.
MSRGC-Net achieves state-of-the-art clustering performance with drastically reduced computational overhead by eliminating the need for iterative training.
Time-series clustering remains challenging due to the inherent trade-off between clustering effectiveness and computational efficiency. Similarity-based methods often suffer from quadratic complexity caused by pairwise distance computations, while deep learning-based approaches typically rely on costly iterative training and a large number of trainable parameters. In this paper, we propose MSRGC-Net, an efficient time-series clustering framework that integrates multiscale reservoir computing, granular-ball-based anchoring graph construction, and consensus learning. MSRGC-Net adopts a training-free reservoir computing paradigm to extract multiscale temporal representations from raw time series without backpropagation, significantly reducing computational overhead. To capture the intrinsic structure of the resulting representations, granular-ball computing is employed to adaptively model data distributions via density-consistent regions, yielding compact and robust anchor graph representations. Furthermore, a consensus-based anchoring graph optimization strategy is introduced to effectively align multiscale reservoir representations and integrate complementary information across temporal scales. Extensive experiments on widely used univariate and multivariate benchmark datasets demonstrate that MSRGC-Net consistently outperforms state-of-the-art methods in clustering performance while maintaining superior computational efficiency.
