Spectra-Guided Neural Tucker Factorization
About
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.
Fusheng Wang, Yikai Hou• 2026
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Tensor completion | NYCTaxi D1 (10:90) | MAE3.4367 | 4 | |
| Tensor completion | NYCTaxi D2 (20:80) | MAE3.1962 | 4 | |
| Tensor completion | PCTemp D3 (10:90) | MAE0.5714 | 4 | |
| Tensor completion | PCTemp D4 (20:80) | MAE0.4744 | 4 | |
| Tensor completion | WSDream QoS-RT D5 (2:98) | MAE0.6487 | 4 | |
| Tensor completion | WSDream QoS-TH D6 (2:98) | MAE4.3689 | 4 |
Showing 6 of 6 rows