Spectra-Guided Neural Tucker Factorization

2026-05-30

AI Summary

This work addresses the challenge of tensor completion for high-dimensional incomplete (HDI) data by proposing a novel approach that integrates continuous spectral embedding with a spatio-temporal collaborative gating (STCG) mechanism. By mapping scalar timestamps into a continuous spectral space, the method effectively captures temporal periodicity, overcoming the limitations of conventional discrete representations. The STCG mechanism explicitly models and filters latent interactions within spatio-temporal contexts, enhancing contextual awareness. Coupled with neural Tucker decomposition, the proposed framework achieves high-accuracy completion across multiple real-world HDI tensor datasets, demonstrating superior parameter efficiency and scalability.

Abstract

This paper proposes Spectra-Guided Neural Tucker Factorization (SG-NTF) for High-Dimensional and Incomplete (HDI) tensor completion. Circumventing discrete representational limits, SG-NTF maps scalar timestamps into a continuous spectral space to abstract temporal periodicities. Concurrently, a Spatio-Temporal Co-Gating (STCG) mechanism explicitly filters latent interactions via multiplicative modulation on spatiotemporal contexts. Evaluations on real-world HDI tensors verify that SG-NTF maintains competitive completion accuracy with parameter efficiency.

Problem

• tensor completion
• high-dimensional
• incomplete data
• temporal periodicities
• spatio-temporal

Innovation

• Spectra-Guided Neural Tucker Factorization
• Tensor Completion
• Spatio-Temporal Co-Gating
• Continuous Spectral Embedding
• High-Dimensional Incomplete Tensors