Research map / Probabilistic and scientific modeling / Molecules and proteins
An Asymmetric Latent Factorization-of-Tensors Model for Relation Analysis
Weiling Li, Zhaoheng Shi, Jiajia Mi, Zhigang Liu, Jialiang Wang, Xin Luo
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use aligned-bipartite constraint on time modes (ALFT) instead of treating all mode pairs as bipartite (FCTN) | LFT model · D2.1 (Guangzhou) | RMSE: 3.613 -> 3.423 | Improved |
| use ALFT constraints versus SVDinsTN's constraint placement | LFT model · D2.1 (Guangzhou) | RMSE: 3.737 -> 3.423 | Improved |
| use ALFT constraint placement instead of Model-3 placement (ablation) | ALFT · D2.1 (Guangzhou) | RMSE: 3.697 -> 3.423 | Improved |
| use ALFT placement instead of Model-1 (ablation) | ALFT · D2.1 (Guangzhou) | RMSE: 3.550 -> 3.423 | Improved |
| activate all diagonal matrices when all modes are entity graphs (ablation) | ATN · D7 (video) | RMSE: 26.576 -> 25.555 | Improved |
| activate diagonal matrices when few or no modes are entity graphs (ablation) | ATN · D4 (UMLS) | RMSE: 0.248 -> 0.365 | Hurt |
| keep time-correlated modes separate rather than merging them (ALFT vs NCF on merged matrix) (ablation) | ALFT · Shanghai traffic (modes merged, Exp 1.2) | RMSE: 6.12 -> 5.54 | Improved |
| remove 75% of mode 2 (entity graph) entities (ablation) | ALFT vs FCTN · Shanghai traffic | RMSE: 5.57 -> 6.15 | Hurt |
| remove 50% of mode 2 entities (ablation) | ALFT vs FCTN · Shanghai traffic | RMSE: 5.57 vs 5.81 | Mixed |
| increase data density from D1.1 to D1.4 (ablation) | ALFT · Shanghai traffic | RMSE: 6.592 -> 5.536 | Improved |
| use ALFT over DAIN | ALFT · D1.1 (Shanghai) | MAE: 3.984 vs 3.930 | No clear effect |
| use ALFT over FCTN on denser data | ALFT · D3.2 (Aarhus) | RMSE: 2.733 -> 2.731 | No clear effect |
Closest papers by meaning
- Determining the Number of Latent Factors in Statistical Multi-Relational Learning (JMLR 2019 · similarity 0.87)
- A tensor factorization model of multilayer network interdependence (JMLR 2024 · similarity 0.83)
- Variational Estimators of the Degree-corrected Latent Block Model for Bipartite Networks (JMLR 2024 · similarity 0.83)
- ER: Equivariance Regularizer for Knowledge Graph Completion (AAAI 2022 · similarity 0.83)
- The Structure of Relation Decoding Linear Operators in Large Language Models (NeurIPS 2025 · similarity 0.82)
- Learning Discrete Latent Variable Structures with Tensor Rank Conditions (NeurIPS 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
