Deep time-series clustering via latent representation alignment

Sangho Lee, Chihyeon Choi, Youngdoo Son

Abstract: Highlights•A novel deep time-series clustering method with topological information is proposed.•Topological information helps align the samples with similar temporal structures.•Our loss function exploits topological information obtained by eigendecomposition.•The proposed method outperforms the state of the arts in deep time-series clustering.