Research map / Label-efficient and robust learning / Domain adaptation and generalization
Deep Transfer Tensor Decomposition with Orthogonal Constraint for Recommender Systems
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Add cross-domain transfer through the Tucker core, side-information learning, and CDAA alignment | DTTD · RB → TA-12M and RB → TA-20M, 60% and 80% training | RMSE: 0.936 -> 0.886; 0.922 -> 0.876; 1.040 -> 0.892; 0.962 -> 0.880 | Helped |
| Add cross-domain transfer through the Tucker core, side-information learning, and CDAA alignment | DTTD · RB → TA-12M and RB → TA-20M, 60% and 80% training | MAE: 0.677 -> 0.621; 0.663 -> 0.608; 0.716 -> 0.666; 0.690 -> 0.662 | Helped |
| Remove cross-domain knowledge transfer (ablation) | DTTD · Three recommendation dataset pairs | RMSE | Hurt |
| Remove side-information autoencoders (ablation) | DTTD · Three recommendation dataset pairs, especially TA datasets | RMSE | Hurt |
| Remove CDAA alignment (ablation) | DTTD · Three recommendation dataset pairs | RMSE | Hurt |
| Replace OC-SDAE with SDAE (ablation) | DTTD · Three recommendation dataset pairs | RMSE | Hurt |
| Remove CDAA after replacing OC-SDAE with SDAE (ablation) | DTTD · Three recommendation dataset pairs | RMSE | Hurt |
| Replace Tucker decomposition with matrix factorization (ablation) | DTTD · Three recommendation dataset pairs | RMSE | Hurt |
| Increase the number of recommended items K | DTTD · Top-K recommendation, K from 2 to 10 | HR@K | Helped |
| Increase the number of recommended items K | DTTD · Top-K recommendation, K from 2 to 10 | NDCG@K | Helped |
| Increase the weight on side information beyond its useful range | DTTD · Parameter analysis | NDCG | Mixed |
| Increase latent-factor dimension | DTTD · Parameter analysis | NDCG | Hurt |
Closest papers by meaning
- A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems (AAAI 2017 · similarity 0.84)
- Over-parameterized Student Model via Tensor Decomposition Boosted Knowledge Distillation (NeurIPS 2024 · similarity 0.82)
- Graph Disentangled Contrastive Learning with Personalized Transfer for Cross-Domain Recommendation (AAAI 2024 · similarity 0.82)
- DR-VAE: Debiased and Representation-enhanced Variational Autoencoder for Collaborative Recommendation (AAAI 2025 · similarity 0.82)
- Geometry-aware training of factorized layers in tensor Tucker format (NeurIPS 2024 · similarity 0.82)
- Efficient Leverage Score Sampling for Tensor Train Decomposition (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.